"Fossies" - the Fresh Open Source Software Archive 
Member "mpr-2.8/malloc.c" (18 Dec 2003, 183958 Bytes) of package /linux/misc/old/mpr-2.8.tar.gz:
As a special service "Fossies" has tried to format the requested source page into HTML format using (guessed) C and C++ source code syntax highlighting (style:
standard) with prefixed line numbers and
code folding option.
Alternatively you can here
view or
download the uninterpreted source code file.
1 /*
2 This is a version (aka dlmalloc) of malloc/free/realloc written by
3 Doug Lea and released to the public domain. Use, modify, and
4 redistribute this code without permission or acknowledgement in any
5 way you wish. Send questions, comments, complaints, performance
6 data, etc to dl@cs.oswego.edu
7
8 * VERSION 2.7.2 Sat Aug 17 09:07:30 2002 Doug Lea (dl at gee)
9
10 Note: There may be an updated version of this malloc obtainable at
11 ftp://gee.cs.oswego.edu/pub/misc/malloc.c
12 Check before installing!
13
14 * Quickstart
15
16 This library is all in one file to simplify the most common usage:
17 ftp it, compile it (-O), and link it into another program. All
18 of the compile-time options default to reasonable values for use on
19 most unix platforms. Compile -DWIN32 for reasonable defaults on windows.
20 You might later want to step through various compile-time and dynamic
21 tuning options.
22
23 For convenience, an include file for code using this malloc is at:
24 ftp://gee.cs.oswego.edu/pub/misc/malloc-2.7.1.h
25 You don't really need this .h file unless you call functions not
26 defined in your system include files. The .h file contains only the
27 excerpts from this file needed for using this malloc on ANSI C/C++
28 systems, so long as you haven't changed compile-time options about
29 naming and tuning parameters. If you do, then you can create your
30 own malloc.h that does include all settings by cutting at the point
31 indicated below.
32
33 * Why use this malloc?
34
35 This is not the fastest, most space-conserving, most portable, or
36 most tunable malloc ever written. However it is among the fastest
37 while also being among the most space-conserving, portable and tunable.
38 Consistent balance across these factors results in a good general-purpose
39 allocator for malloc-intensive programs.
40
41 The main properties of the algorithms are:
42 * For large (>= 512 bytes) requests, it is a pure best-fit allocator,
43 with ties normally decided via FIFO (i.e. least recently used).
44 * For small (<= 64 bytes by default) requests, it is a caching
45 allocator, that maintains pools of quickly recycled chunks.
46 * In between, and for combinations of large and small requests, it does
47 the best it can trying to meet both goals at once.
48 * For very large requests (>= 128KB by default), it relies on system
49 memory mapping facilities, if supported.
50
51 For a longer but slightly out of date high-level description, see
52 http://gee.cs.oswego.edu/dl/html/malloc.html
53
54 You may already by default be using a C library containing a malloc
55 that is based on some version of this malloc (for example in
56 linux). You might still want to use the one in this file in order to
57 customize settings or to avoid overheads associated with library
58 versions.
59
60 * Contents, described in more detail in "description of public routines" below.
61
62 Standard (ANSI/SVID/...) functions:
63 malloc(size_t n);
64 calloc(size_t n_elements, size_t element_size);
65 free(Void_t* p);
66 realloc(Void_t* p, size_t n);
67 memalign(size_t alignment, size_t n);
68 valloc(size_t n);
69 mallinfo()
70 mallopt(int parameter_number, int parameter_value)
71
72 Additional functions:
73 independent_calloc(size_t n_elements, size_t size, Void_t* chunks[]);
74 independent_comalloc(size_t n_elements, size_t sizes[], Void_t* chunks[]);
75 pvalloc(size_t n);
76 cfree(Void_t* p);
77 malloc_trim(size_t pad);
78 malloc_usable_size(Void_t* p);
79 malloc_stats();
80
81 * Vital statistics:
82
83 Supported pointer representation: 4 or 8 bytes
84 Supported size_t representation: 4 or 8 bytes
85 Note that size_t is allowed to be 4 bytes even if pointers are 8.
86 You can adjust this by defining INTERNAL_SIZE_T
87
88 Alignment: 2 * sizeof(size_t) (default)
89 (i.e., 8 byte alignment with 4byte size_t). This suffices for
90 nearly all current machines and C compilers. However, you can
91 define MALLOC_ALIGNMENT to be wider than this if necessary.
92
93 Minimum overhead per allocated chunk: 4 or 8 bytes
94 Each malloced chunk has a hidden word of overhead holding size
95 and status information.
96
97 Minimum allocated size: 4-byte ptrs: 16 bytes (including 4 overhead)
98 8-byte ptrs: 24/32 bytes (including, 4/8 overhead)
99
100 When a chunk is freed, 12 (for 4byte ptrs) or 20 (for 8 byte
101 ptrs but 4 byte size) or 24 (for 8/8) additional bytes are
102 needed; 4 (8) for a trailing size field and 8 (16) bytes for
103 free list pointers. Thus, the minimum allocatable size is
104 16/24/32 bytes.
105
106 Even a request for zero bytes (i.e., malloc(0)) returns a
107 pointer to something of the minimum allocatable size.
108
109 The maximum overhead wastage (i.e., number of extra bytes
110 allocated than were requested in malloc) is less than or equal
111 to the minimum size, except for requests >= mmap_threshold that
112 are serviced via mmap(), where the worst case wastage is 2 *
113 sizeof(size_t) bytes plus the remainder from a system page (the
114 minimal mmap unit); typically 4096 or 8192 bytes.
115
116 Maximum allocated size: 4-byte size_t: 2^32 minus about two pages
117 8-byte size_t: 2^64 minus about two pages
118
119 It is assumed that (possibly signed) size_t values suffice to
120 represent chunk sizes. `Possibly signed' is due to the fact
121 that `size_t' may be defined on a system as either a signed or
122 an unsigned type. The ISO C standard says that it must be
123 unsigned, but a few systems are known not to adhere to this.
124 Additionally, even when size_t is unsigned, sbrk (which is by
125 default used to obtain memory from system) accepts signed
126 arguments, and may not be able to handle size_t-wide arguments
127 with negative sign bit. Generally, values that would
128 appear as negative after accounting for overhead and alignment
129 are supported only via mmap(), which does not have this
130 limitation.
131
132 Requests for sizes outside the allowed range will perform an optional
133 failure action and then return null. (Requests may also
134 also fail because a system is out of memory.)
135
136 Thread-safety: NOT thread-safe unless USE_MALLOC_LOCK defined
137
138 When USE_MALLOC_LOCK is defined, wrappers are created to
139 surround every public call with either a pthread mutex or
140 a win32 spinlock (depending on WIN32). This is not
141 especially fast, and can be a major bottleneck.
142 It is designed only to provide minimal protection
143 in concurrent environments, and to provide a basis for
144 extensions. If you are using malloc in a concurrent program,
145 you would be far better off obtaining ptmalloc, which is
146 derived from a version of this malloc, and is well-tuned for
147 concurrent programs. (See http://www.malloc.de) Note that
148 even when USE_MALLOC_LOCK is defined, you can can guarantee
149 full thread-safety only if no threads acquire memory through
150 direct calls to MORECORE or other system-level allocators.
151
152 Compliance: I believe it is compliant with the 1997 Single Unix Specification
153 (See http://www.opennc.org). Also SVID/XPG, ANSI C, and probably
154 others as well.
155
156 * Synopsis of compile-time options:
157
158 People have reported using previous versions of this malloc on all
159 versions of Unix, sometimes by tweaking some of the defines
160 below. It has been tested most extensively on Solaris and
161 Linux. It is also reported to work on WIN32 platforms.
162 People also report using it in stand-alone embedded systems.
163
164 The implementation is in straight, hand-tuned ANSI C. It is not
165 at all modular. (Sorry!) It uses a lot of macros. To be at all
166 usable, this code should be compiled using an optimizing compiler
167 (for example gcc -O3) that can simplify expressions and control
168 paths. (FAQ: some macros import variables as arguments rather than
169 declare locals because people reported that some debuggers
170 otherwise get confused.)
171
172 OPTION DEFAULT VALUE
173
174 Compilation Environment options:
175
176 __STD_C derived from C compiler defines
177 WIN32 NOT defined
178 HAVE_MEMCPY defined
179 USE_MEMCPY 1 if HAVE_MEMCPY is defined
180 HAVE_MMAP defined as 1
181 MMAP_CLEARS 1
182 HAVE_MREMAP 0 unless linux defined
183 malloc_getpagesize derived from system #includes, or 4096 if not
184 HAVE_USR_INCLUDE_MALLOC_H NOT defined
185 LACKS_UNISTD_H NOT defined unless WIN32
186 LACKS_SYS_PARAM_H NOT defined unless WIN32
187 LACKS_SYS_MMAN_H NOT defined unless WIN32
188 LACKS_FCNTL_H NOT defined
189
190 Changing default word sizes:
191
192 INTERNAL_SIZE_T size_t
193 MALLOC_ALIGNMENT 2 * sizeof(INTERNAL_SIZE_T)
194 PTR_UINT unsigned long
195 CHUNK_SIZE_T unsigned long
196
197 Configuration and functionality options:
198
199 USE_DL_PREFIX NOT defined
200 USE_PUBLIC_MALLOC_WRAPPERS NOT defined
201 USE_MALLOC_LOCK NOT defined
202 DEBUG NOT defined
203 REALLOC_ZERO_BYTES_FREES NOT defined
204 MALLOC_FAILURE_ACTION errno = ENOMEM, if __STD_C defined, else no-op
205 TRIM_FASTBINS 0
206 FIRST_SORTED_BIN_SIZE 512
207
208 Options for customizing MORECORE:
209
210 MORECORE sbrk
211 MORECORE_CONTIGUOUS 1
212 MORECORE_CANNOT_TRIM NOT defined
213 MMAP_AS_MORECORE_SIZE (1024 * 1024)
214
215 Tuning options that are also dynamically changeable via mallopt:
216
217 DEFAULT_MXFAST 64
218 DEFAULT_TRIM_THRESHOLD 256 * 1024
219 DEFAULT_TOP_PAD 0
220 DEFAULT_MMAP_THRESHOLD 256 * 1024
221 DEFAULT_MMAP_MAX 65536
222
223 There are several other #defined constants and macros that you
224 probably don't want to touch unless you are extending or adapting malloc.
225 */
226
227 /*
228 WIN32 sets up defaults for MS environment and compilers.
229 Otherwise defaults are for unix.
230 */
231
232 /* #define WIN32 */
233
234 #ifdef WIN32
235
236 #define WIN32_LEAN_AND_MEAN
237 #include <windows.h>
238
239 /* Win32 doesn't supply or need the following headers */
240 #define LACKS_UNISTD_H
241 #define LACKS_SYS_PARAM_H
242 #define LACKS_SYS_MMAN_H
243
244 /* Use the supplied emulation of sbrk */
245 #define MORECORE sbrk
246 #define MORECORE_CONTIGUOUS 1
247 #define MORECORE_FAILURE ((void*)(-1))
248
249 /* Use the supplied emulation of mmap and munmap */
250 #define HAVE_MMAP 1
251 #define MUNMAP_FAILURE (-1)
252 #define MMAP_CLEARS 1
253
254 /* These values don't really matter in windows mmap emulation */
255 #define MAP_PRIVATE 1
256 #define MAP_ANONYMOUS 2
257 #define PROT_READ 1
258 #define PROT_WRITE 2
259
260 /* Emulation functions defined at the end of this file */
261
262 /* If USE_MALLOC_LOCK, use supplied critical-section-based lock functions */
263 #ifdef USE_MALLOC_LOCK
264 static int slwait(int *sl);
265 static int slrelease(int *sl);
266 #endif
267
268 static long getpagesize(void);
269 static long getregionsize(void);
270 static void *sbrk(long size);
271 static void *mmap(void *ptr, long size, long prot, long type, long handle, long arg);
272 static long munmap(void *ptr, long size);
273
274 static void vminfo (unsigned long*free, unsigned long*reserved, unsigned long*committed);
275 static int cpuinfo (int whole, unsigned long*kernel, unsigned long*user);
276
277 #endif
278
279 /*
280 __STD_C should be nonzero if using ANSI-standard C compiler, a C++
281 compiler, or a C compiler sufficiently close to ANSI to get away
282 with it.
283 */
284
285 #ifndef __STD_C
286 #if defined(__STDC__) || defined(_cplusplus)
287 #define __STD_C 1
288 #else
289 #define __STD_C 0
290 #endif
291 #endif /*__STD_C*/
292
293
294 /*
295 Void_t* is the pointer type that malloc should say it returns
296 */
297
298 #ifndef Void_t
299 #if (__STD_C || defined(WIN32))
300 #define Void_t void
301 #else
302 #define Void_t char
303 #endif
304 #endif /*Void_t*/
305
306 #if __STD_C
307 #include <stddef.h> /* for size_t */
308 #else
309 #include <sys/types.h>
310 #endif
311
312 #ifdef __cplusplus
313 extern "C" {
314 #endif
315
316 /* define LACKS_UNISTD_H if your system does not have a <unistd.h>. */
317
318 /* #define LACKS_UNISTD_H */
319
320 #ifndef LACKS_UNISTD_H
321 #include <unistd.h>
322 #endif
323
324 /* define LACKS_SYS_PARAM_H if your system does not have a <sys/param.h>. */
325
326 /* #define LACKS_SYS_PARAM_H */
327
328
329 #include <stdio.h> /* needed for malloc_stats */
330 #include <errno.h> /* needed for optional MALLOC_FAILURE_ACTION */
331
332
333 /*
334 Debugging:
335
336 Because freed chunks may be overwritten with bookkeeping fields, this
337 malloc will often die when freed memory is overwritten by user
338 programs. This can be very effective (albeit in an annoying way)
339 in helping track down dangling pointers.
340
341 If you compile with -DDEBUG, a number of assertion checks are
342 enabled that will catch more memory errors. You probably won't be
343 able to make much sense of the actual assertion errors, but they
344 should help you locate incorrectly overwritten memory. The
345 checking is fairly extensive, and will slow down execution
346 noticeably. Calling malloc_stats or mallinfo with DEBUG set will
347 attempt to check every non-mmapped allocated and free chunk in the
348 course of computing the summmaries. (By nature, mmapped regions
349 cannot be checked very much automatically.)
350
351 Setting DEBUG may also be helpful if you are trying to modify
352 this code. The assertions in the check routines spell out in more
353 detail the assumptions and invariants underlying the algorithms.
354
355 Setting DEBUG does NOT provide an automated mechanism for checking
356 that all accesses to malloced memory stay within their
357 bounds. However, there are several add-ons and adaptations of this
358 or other mallocs available that do this.
359 */
360
361 #if DEBUG
362 #include <assert.h>
363 #else
364 #define assert(x) ((void)0)
365 #endif
366
367 /*
368 The unsigned integer type used for comparing any two chunk sizes.
369 This should be at least as wide as size_t, but should not be signed.
370 */
371
372 #ifndef CHUNK_SIZE_T
373 #define CHUNK_SIZE_T unsigned long
374 #endif
375
376 /*
377 The unsigned integer type used to hold addresses when they are are
378 manipulated as integers. Except that it is not defined on all
379 systems, intptr_t would suffice.
380 */
381 #ifndef PTR_UINT
382 #define PTR_UINT unsigned long
383 #endif
384
385
386 /*
387 INTERNAL_SIZE_T is the word-size used for internal bookkeeping
388 of chunk sizes.
389
390 The default version is the same as size_t.
391
392 While not strictly necessary, it is best to define this as an
393 unsigned type, even if size_t is a signed type. This may avoid some
394 artificial size limitations on some systems.
395
396 On a 64-bit machine, you may be able to reduce malloc overhead by
397 defining INTERNAL_SIZE_T to be a 32 bit `unsigned int' at the
398 expense of not being able to handle more than 2^32 of malloced
399 space. If this limitation is acceptable, you are encouraged to set
400 this unless you are on a platform requiring 16byte alignments. In
401 this case the alignment requirements turn out to negate any
402 potential advantages of decreasing size_t word size.
403
404 Implementors: Beware of the possible combinations of:
405 - INTERNAL_SIZE_T might be signed or unsigned, might be 32 or 64 bits,
406 and might be the same width as int or as long
407 - size_t might have different width and signedness as INTERNAL_SIZE_T
408 - int and long might be 32 or 64 bits, and might be the same width
409 To deal with this, most comparisons and difference computations
410 among INTERNAL_SIZE_Ts should cast them to CHUNK_SIZE_T, being
411 aware of the fact that casting an unsigned int to a wider long does
412 not sign-extend. (This also makes checking for negative numbers
413 awkward.) Some of these casts result in harmless compiler warnings
414 on some systems.
415 */
416
417 #ifndef INTERNAL_SIZE_T
418 #define INTERNAL_SIZE_T size_t
419 #endif
420
421 /* The corresponding word size */
422 #define SIZE_SZ (sizeof(INTERNAL_SIZE_T))
423
424
425
426 /*
427 MALLOC_ALIGNMENT is the minimum alignment for malloc'ed chunks.
428 It must be a power of two at least 2 * SIZE_SZ, even on machines
429 for which smaller alignments would suffice. It may be defined as
430 larger than this though. Note however that code and data structures
431 are optimized for the case of 8-byte alignment.
432 */
433
434
435 #ifndef MALLOC_ALIGNMENT
436 #define MALLOC_ALIGNMENT (2 * SIZE_SZ)
437 #endif
438
439 /* The corresponding bit mask value */
440 #define MALLOC_ALIGN_MASK (MALLOC_ALIGNMENT - 1)
441
442
443
444 /*
445 REALLOC_ZERO_BYTES_FREES should be set if a call to
446 realloc with zero bytes should be the same as a call to free.
447 Some people think it should. Otherwise, since this malloc
448 returns a unique pointer for malloc(0), so does realloc(p, 0).
449 */
450
451 /* #define REALLOC_ZERO_BYTES_FREES */
452
453 /*
454 TRIM_FASTBINS controls whether free() of a very small chunk can
455 immediately lead to trimming. Setting to true (1) can reduce memory
456 footprint, but will almost always slow down programs that use a lot
457 of small chunks.
458
459 Define this only if you are willing to give up some speed to more
460 aggressively reduce system-level memory footprint when releasing
461 memory in programs that use many small chunks. You can get
462 essentially the same effect by setting MXFAST to 0, but this can
463 lead to even greater slowdowns in programs using many small chunks.
464 TRIM_FASTBINS is an in-between compile-time option, that disables
465 only those chunks bordering topmost memory from being placed in
466 fastbins.
467 */
468
469 #ifndef TRIM_FASTBINS
470 #define TRIM_FASTBINS 0
471 #endif
472
473
474 /*
475 USE_DL_PREFIX will prefix all public routines with the string 'dl'.
476 This is necessary when you only want to use this malloc in one part
477 of a program, using your regular system malloc elsewhere.
478 */
479
480 /* #define USE_DL_PREFIX */
481
482
483 /*
484 USE_MALLOC_LOCK causes wrapper functions to surround each
485 callable routine with pthread mutex lock/unlock.
486
487 USE_MALLOC_LOCK forces USE_PUBLIC_MALLOC_WRAPPERS to be defined
488 */
489
490
491 /* #define USE_MALLOC_LOCK */
492
493
494 /*
495 If USE_PUBLIC_MALLOC_WRAPPERS is defined, every public routine is
496 actually a wrapper function that first calls MALLOC_PREACTION, then
497 calls the internal routine, and follows it with
498 MALLOC_POSTACTION. This is needed for locking, but you can also use
499 this, without USE_MALLOC_LOCK, for purposes of interception,
500 instrumentation, etc. It is a sad fact that using wrappers often
501 noticeably degrades performance of malloc-intensive programs.
502 */
503
504 #ifdef USE_MALLOC_LOCK
505 #define USE_PUBLIC_MALLOC_WRAPPERS
506 #else
507 /* #define USE_PUBLIC_MALLOC_WRAPPERS */
508 #endif
509
510
511 /*
512 Two-phase name translation.
513 All of the actual routines are given mangled names.
514 When wrappers are used, they become the public callable versions.
515 When DL_PREFIX is used, the callable names are prefixed.
516 */
517
518 #ifndef USE_PUBLIC_MALLOC_WRAPPERS
519 #define cALLOc public_cALLOc
520 #define fREe public_fREe
521 #define cFREe public_cFREe
522 #define mALLOc public_mALLOc
523 #define mEMALIGn public_mEMALIGn
524 #define rEALLOc public_rEALLOc
525 #define vALLOc public_vALLOc
526 #define pVALLOc public_pVALLOc
527 #define mALLINFo public_mALLINFo
528 #define mALLOPt public_mALLOPt
529 #define mTRIm public_mTRIm
530 #define mSTATs public_mSTATs
531 #define mUSABLe public_mUSABLe
532 #define iCALLOc public_iCALLOc
533 #define iCOMALLOc public_iCOMALLOc
534 #endif
535
536 #ifdef USE_DL_PREFIX
537 #define public_cALLOc dlcalloc
538 #define public_fREe dlfree
539 #define public_cFREe dlcfree
540 #define public_mALLOc dlmalloc
541 #define public_mEMALIGn dlmemalign
542 #define public_rEALLOc dlrealloc
543 #define public_vALLOc dlvalloc
544 #define public_pVALLOc dlpvalloc
545 #define public_mALLINFo dlmallinfo
546 #define public_mALLOPt dlmallopt
547 #define public_mTRIm dlmalloc_trim
548 #define public_mSTATs dlmalloc_stats
549 #define public_mUSABLe dlmalloc_usable_size
550 #define public_iCALLOc dlindependent_calloc
551 #define public_iCOMALLOc dlindependent_comalloc
552 #else /* USE_DL_PREFIX */
553 #define public_cALLOc calloc
554 #define public_fREe free
555 #define public_cFREe cfree
556 #define public_mALLOc malloc
557 #define public_mEMALIGn memalign
558 #define public_rEALLOc realloc
559 #define public_vALLOc valloc
560 #define public_pVALLOc pvalloc
561 #define public_mALLINFo mallinfo
562 #define public_mALLOPt mallopt
563 #define public_mTRIm malloc_trim
564 #define public_mSTATs malloc_stats
565 #define public_mUSABLe malloc_usable_size
566 #define public_iCALLOc independent_calloc
567 #define public_iCOMALLOc independent_comalloc
568 #endif /* USE_DL_PREFIX */
569
570
571 /*
572 HAVE_MEMCPY should be defined if you are not otherwise using
573 ANSI STD C, but still have memcpy and memset in your C library
574 and want to use them in calloc and realloc. Otherwise simple
575 macro versions are defined below.
576
577 USE_MEMCPY should be defined as 1 if you actually want to
578 have memset and memcpy called. People report that the macro
579 versions are faster than libc versions on some systems.
580
581 Even if USE_MEMCPY is set to 1, loops to copy/clear small chunks
582 (of <= 36 bytes) are manually unrolled in realloc and calloc.
583 */
584
585 #define HAVE_MEMCPY
586
587 #ifndef USE_MEMCPY
588 #ifdef HAVE_MEMCPY
589 #define USE_MEMCPY 1
590 #else
591 #define USE_MEMCPY 0
592 #endif
593 #endif
594
595
596 #if (__STD_C || defined(HAVE_MEMCPY))
597
598 #ifdef WIN32
599 /* On Win32 memset and memcpy are already declared in windows.h */
600 #else
601 #if __STD_C
602 void* memset(void*, int, size_t);
603 void* memcpy(void*, const void*, size_t);
604 #else
605 Void_t* memset();
606 Void_t* memcpy();
607 #endif
608 #endif
609 #endif
610
611 /*
612 MALLOC_FAILURE_ACTION is the action to take before "return 0" when
613 malloc fails to be able to return memory, either because memory is
614 exhausted or because of illegal arguments.
615
616 By default, sets errno if running on STD_C platform, else does nothing.
617 */
618
619 #ifndef MALLOC_FAILURE_ACTION
620 #if __STD_C
621 #define MALLOC_FAILURE_ACTION \
622 errno = ENOMEM;
623
624 #else
625 #define MALLOC_FAILURE_ACTION
626 #endif
627 #endif
628
629 /*
630 MORECORE-related declarations. By default, rely on sbrk
631 */
632
633
634 #ifdef LACKS_UNISTD_H
635 #if !defined(__FreeBSD__) && !defined(__OpenBSD__) && !defined(__NetBSD__)
636 #if __STD_C
637 extern Void_t* sbrk(ptrdiff_t);
638 #else
639 extern Void_t* sbrk();
640 #endif
641 #endif
642 #endif
643
644 /*
645 MORECORE is the name of the routine to call to obtain more memory
646 from the system. See below for general guidance on writing
647 alternative MORECORE functions, as well as a version for WIN32 and a
648 sample version for pre-OSX macos.
649 */
650
651 #ifndef MORECORE
652 #define MORECORE sbrk
653 #endif
654
655 /*
656 MORECORE_FAILURE is the value returned upon failure of MORECORE
657 as well as mmap. Since it cannot be an otherwise valid memory address,
658 and must reflect values of standard sys calls, you probably ought not
659 try to redefine it.
660 */
661
662 #ifndef MORECORE_FAILURE
663 #define MORECORE_FAILURE (-1)
664 #endif
665
666 /*
667 If MORECORE_CONTIGUOUS is true, take advantage of fact that
668 consecutive calls to MORECORE with positive arguments always return
669 contiguous increasing addresses. This is true of unix sbrk. Even
670 if not defined, when regions happen to be contiguous, malloc will
671 permit allocations spanning regions obtained from different
672 calls. But defining this when applicable enables some stronger
673 consistency checks and space efficiencies.
674 */
675
676 #ifndef MORECORE_CONTIGUOUS
677 #define MORECORE_CONTIGUOUS 1
678 #endif
679
680 /*
681 Define MORECORE_CANNOT_TRIM if your version of MORECORE
682 cannot release space back to the system when given negative
683 arguments. This is generally necessary only if you are using
684 a hand-crafted MORECORE function that cannot handle negative arguments.
685 */
686
687 /* #define MORECORE_CANNOT_TRIM */
688
689
690 /*
691 Define HAVE_MMAP as true to optionally make malloc() use mmap() to
692 allocate very large blocks. These will be returned to the
693 operating system immediately after a free(). Also, if mmap
694 is available, it is used as a backup strategy in cases where
695 MORECORE fails to provide space from system.
696
697 This malloc is best tuned to work with mmap for large requests.
698 If you do not have mmap, operations involving very large chunks (1MB
699 or so) may be slower than you'd like.
700 */
701
702 #ifndef HAVE_MMAP
703 #define HAVE_MMAP 1
704 #endif
705
706 #if HAVE_MMAP
707 /*
708 Standard unix mmap using /dev/zero clears memory so calloc doesn't
709 need to.
710 */
711
712 #ifndef MMAP_CLEARS
713 #define MMAP_CLEARS 1
714 #endif
715
716 #else /* no mmap */
717 #ifndef MMAP_CLEARS
718 #define MMAP_CLEARS 0
719 #endif
720 #endif
721
722
723 /*
724 MMAP_AS_MORECORE_SIZE is the minimum mmap size argument to use if
725 sbrk fails, and mmap is used as a backup (which is done only if
726 HAVE_MMAP). The value must be a multiple of page size. This
727 backup strategy generally applies only when systems have "holes" in
728 address space, so sbrk cannot perform contiguous expansion, but
729 there is still space available on system. On systems for which
730 this is known to be useful (i.e. most linux kernels), this occurs
731 only when programs allocate huge amounts of memory. Between this,
732 and the fact that mmap regions tend to be limited, the size should
733 be large, to avoid too many mmap calls and thus avoid running out
734 of kernel resources.
735 */
736
737 #ifndef MMAP_AS_MORECORE_SIZE
738 #define MMAP_AS_MORECORE_SIZE (1024 * 1024)
739 #endif
740
741 /*
742 Define HAVE_MREMAP to make realloc() use mremap() to re-allocate
743 large blocks. This is currently only possible on Linux with
744 kernel versions newer than 1.3.77.
745 */
746
747 #ifndef HAVE_MREMAP
748 #ifdef linux
749 #define HAVE_MREMAP 1
750 #else
751 #define HAVE_MREMAP 0
752 #endif
753
754 #endif /* HAVE_MMAP */
755
756
757 /*
758 The system page size. To the extent possible, this malloc manages
759 memory from the system in page-size units. Note that this value is
760 cached during initialization into a field of malloc_state. So even
761 if malloc_getpagesize is a function, it is only called once.
762
763 The following mechanics for getpagesize were adapted from bsd/gnu
764 getpagesize.h. If none of the system-probes here apply, a value of
765 4096 is used, which should be OK: If they don't apply, then using
766 the actual value probably doesn't impact performance.
767 */
768
769
770 #ifndef malloc_getpagesize
771
772 #ifndef LACKS_UNISTD_H
773 # include <unistd.h>
774 #endif
775
776 # ifdef _SC_PAGESIZE /* some SVR4 systems omit an underscore */
777 # ifndef _SC_PAGE_SIZE
778 # define _SC_PAGE_SIZE _SC_PAGESIZE
779 # endif
780 # endif
781
782 # ifdef _SC_PAGE_SIZE
783 # define malloc_getpagesize sysconf(_SC_PAGE_SIZE)
784 # else
785 # if defined(BSD) || defined(DGUX) || defined(HAVE_GETPAGESIZE)
786 extern size_t getpagesize();
787 # define malloc_getpagesize getpagesize()
788 # else
789 # ifdef WIN32 /* use supplied emulation of getpagesize */
790 # define malloc_getpagesize getpagesize()
791 # else
792 # ifndef LACKS_SYS_PARAM_H
793 # include <sys/param.h>
794 # endif
795 # ifdef EXEC_PAGESIZE
796 # define malloc_getpagesize EXEC_PAGESIZE
797 # else
798 # ifdef NBPG
799 # ifndef CLSIZE
800 # define malloc_getpagesize NBPG
801 # else
802 # define malloc_getpagesize (NBPG * CLSIZE)
803 # endif
804 # else
805 # ifdef NBPC
806 # define malloc_getpagesize NBPC
807 # else
808 # ifdef PAGESIZE
809 # define malloc_getpagesize PAGESIZE
810 # else /* just guess */
811 # define malloc_getpagesize (4096)
812 # endif
813 # endif
814 # endif
815 # endif
816 # endif
817 # endif
818 # endif
819 #endif
820
821 /*
822 This version of malloc supports the standard SVID/XPG mallinfo
823 routine that returns a struct containing usage properties and
824 statistics. It should work on any SVID/XPG compliant system that has
825 a /usr/include/malloc.h defining struct mallinfo. (If you'd like to
826 install such a thing yourself, cut out the preliminary declarations
827 as described above and below and save them in a malloc.h file. But
828 there's no compelling reason to bother to do this.)
829
830 The main declaration needed is the mallinfo struct that is returned
831 (by-copy) by mallinfo(). The SVID/XPG malloinfo struct contains a
832 bunch of fields that are not even meaningful in this version of
833 malloc. These fields are are instead filled by mallinfo() with
834 other numbers that might be of interest.
835
836 HAVE_USR_INCLUDE_MALLOC_H should be set if you have a
837 /usr/include/malloc.h file that includes a declaration of struct
838 mallinfo. If so, it is included; else an SVID2/XPG2 compliant
839 version is declared below. These must be precisely the same for
840 mallinfo() to work. The original SVID version of this struct,
841 defined on most systems with mallinfo, declares all fields as
842 ints. But some others define as unsigned long. If your system
843 defines the fields using a type of different width than listed here,
844 you must #include your system version and #define
845 HAVE_USR_INCLUDE_MALLOC_H.
846 */
847
848 /* #define HAVE_USR_INCLUDE_MALLOC_H */
849
850 #ifdef HAVE_USR_INCLUDE_MALLOC_H
851 #include "/usr/include/malloc.h"
852 #else
853
854 /* SVID2/XPG mallinfo structure */
855
856 struct mallinfo {
857 int arena; /* non-mmapped space allocated from system */
858 int ordblks; /* number of free chunks */
859 int smblks; /* number of fastbin blocks */
860 int hblks; /* number of mmapped regions */
861 int hblkhd; /* space in mmapped regions */
862 int usmblks; /* maximum total allocated space */
863 int fsmblks; /* space available in freed fastbin blocks */
864 int uordblks; /* total allocated space */
865 int fordblks; /* total free space */
866 int keepcost; /* top-most, releasable (via malloc_trim) space */
867 };
868
869 /*
870 SVID/XPG defines four standard parameter numbers for mallopt,
871 normally defined in malloc.h. Only one of these (M_MXFAST) is used
872 in this malloc. The others (M_NLBLKS, M_GRAIN, M_KEEP) don't apply,
873 so setting them has no effect. But this malloc also supports other
874 options in mallopt described below.
875 */
876 #endif
877
878
879 /* ---------- description of public routines ------------ */
880
881 /*
882 malloc(size_t n)
883 Returns a pointer to a newly allocated chunk of at least n bytes, or null
884 if no space is available. Additionally, on failure, errno is
885 set to ENOMEM on ANSI C systems.
886
887 If n is zero, malloc returns a minumum-sized chunk. (The minimum
888 size is 16 bytes on most 32bit systems, and 24 or 32 bytes on 64bit
889 systems.) On most systems, size_t is an unsigned type, so calls
890 with negative arguments are interpreted as requests for huge amounts
891 of space, which will often fail. The maximum supported value of n
892 differs across systems, but is in all cases less than the maximum
893 representable value of a size_t.
894 */
895 #if __STD_C
896 Void_t* public_mALLOc(size_t);
897 #else
898 Void_t* public_mALLOc();
899 #endif
900
901 /*
902 free(Void_t* p)
903 Releases the chunk of memory pointed to by p, that had been previously
904 allocated using malloc or a related routine such as realloc.
905 It has no effect if p is null. It can have arbitrary (i.e., bad!)
906 effects if p has already been freed.
907
908 Unless disabled (using mallopt), freeing very large spaces will
909 when possible, automatically trigger operations that give
910 back unused memory to the system, thus reducing program footprint.
911 */
912 #if __STD_C
913 void public_fREe(Void_t*);
914 #else
915 void public_fREe();
916 #endif
917
918 /*
919 calloc(size_t n_elements, size_t element_size);
920 Returns a pointer to n_elements * element_size bytes, with all locations
921 set to zero.
922 */
923 #if __STD_C
924 Void_t* public_cALLOc(size_t, size_t);
925 #else
926 Void_t* public_cALLOc();
927 #endif
928
929 /*
930 realloc(Void_t* p, size_t n)
931 Returns a pointer to a chunk of size n that contains the same data
932 as does chunk p up to the minimum of (n, p's size) bytes, or null
933 if no space is available.
934
935 The returned pointer may or may not be the same as p. The algorithm
936 prefers extending p when possible, otherwise it employs the
937 equivalent of a malloc-copy-free sequence.
938
939 If p is null, realloc is equivalent to malloc.
940
941 If space is not available, realloc returns null, errno is set (if on
942 ANSI) and p is NOT freed.
943
944 if n is for fewer bytes than already held by p, the newly unused
945 space is lopped off and freed if possible. Unless the #define
946 REALLOC_ZERO_BYTES_FREES is set, realloc with a size argument of
947 zero (re)allocates a minimum-sized chunk.
948
949 Large chunks that were internally obtained via mmap will always
950 be reallocated using malloc-copy-free sequences unless
951 the system supports MREMAP (currently only linux).
952
953 The old unix realloc convention of allowing the last-free'd chunk
954 to be used as an argument to realloc is not supported.
955 */
956 #if __STD_C
957 Void_t* public_rEALLOc(Void_t*, size_t);
958 #else
959 Void_t* public_rEALLOc();
960 #endif
961
962 /*
963 memalign(size_t alignment, size_t n);
964 Returns a pointer to a newly allocated chunk of n bytes, aligned
965 in accord with the alignment argument.
966
967 The alignment argument should be a power of two. If the argument is
968 not a power of two, the nearest greater power is used.
969 8-byte alignment is guaranteed by normal malloc calls, so don't
970 bother calling memalign with an argument of 8 or less.
971
972 Overreliance on memalign is a sure way to fragment space.
973 */
974 #if __STD_C
975 Void_t* public_mEMALIGn(size_t, size_t);
976 #else
977 Void_t* public_mEMALIGn();
978 #endif
979
980 /*
981 valloc(size_t n);
982 Equivalent to memalign(pagesize, n), where pagesize is the page
983 size of the system. If the pagesize is unknown, 4096 is used.
984 */
985 #if __STD_C
986 Void_t* public_vALLOc(size_t);
987 #else
988 Void_t* public_vALLOc();
989 #endif
990
991
992
993 /*
994 mallopt(int parameter_number, int parameter_value)
995 Sets tunable parameters The format is to provide a
996 (parameter-number, parameter-value) pair. mallopt then sets the
997 corresponding parameter to the argument value if it can (i.e., so
998 long as the value is meaningful), and returns 1 if successful else
999 0. SVID/XPG/ANSI defines four standard param numbers for mallopt,
1000 normally defined in malloc.h. Only one of these (M_MXFAST) is used
1001 in this malloc. The others (M_NLBLKS, M_GRAIN, M_KEEP) don't apply,
1002 so setting them has no effect. But this malloc also supports four
1003 other options in mallopt. See below for details. Briefly, supported
1004 parameters are as follows (listed defaults are for "typical"
1005 configurations).
1006
1007 Symbol param # default allowed param values
1008 M_MXFAST 1 64 0-80 (0 disables fastbins)
1009 M_TRIM_THRESHOLD -1 256*1024 any (-1U disables trimming)
1010 M_TOP_PAD -2 0 any
1011 M_MMAP_THRESHOLD -3 256*1024 any (or 0 if no MMAP support)
1012 M_MMAP_MAX -4 65536 any (0 disables use of mmap)
1013 */
1014 #if __STD_C
1015 int public_mALLOPt(int, int);
1016 #else
1017 int public_mALLOPt();
1018 #endif
1019
1020
1021 /*
1022 mallinfo()
1023 Returns (by copy) a struct containing various summary statistics:
1024
1025 arena: current total non-mmapped bytes allocated from system
1026 ordblks: the number of free chunks
1027 smblks: the number of fastbin blocks (i.e., small chunks that
1028 have been freed but not use resused or consolidated)
1029 hblks: current number of mmapped regions
1030 hblkhd: total bytes held in mmapped regions
1031 usmblks: the maximum total allocated space. This will be greater
1032 than current total if trimming has occurred.
1033 fsmblks: total bytes held in fastbin blocks
1034 uordblks: current total allocated space (normal or mmapped)
1035 fordblks: total free space
1036 keepcost: the maximum number of bytes that could ideally be released
1037 back to system via malloc_trim. ("ideally" means that
1038 it ignores page restrictions etc.)
1039
1040 Because these fields are ints, but internal bookkeeping may
1041 be kept as longs, the reported values may wrap around zero and
1042 thus be inaccurate.
1043 */
1044 #if __STD_C
1045 struct mallinfo public_mALLINFo(void);
1046 #else
1047 struct mallinfo public_mALLINFo();
1048 #endif
1049
1050 /*
1051 independent_calloc(size_t n_elements, size_t element_size, Void_t* chunks[]);
1052
1053 independent_calloc is similar to calloc, but instead of returning a
1054 single cleared space, it returns an array of pointers to n_elements
1055 independent elements that can hold contents of size elem_size, each
1056 of which starts out cleared, and can be independently freed,
1057 realloc'ed etc. The elements are guaranteed to be adjacently
1058 allocated (this is not guaranteed to occur with multiple callocs or
1059 mallocs), which may also improve cache locality in some
1060 applications.
1061
1062 The "chunks" argument is optional (i.e., may be null, which is
1063 probably the most typical usage). If it is null, the returned array
1064 is itself dynamically allocated and should also be freed when it is
1065 no longer needed. Otherwise, the chunks array must be of at least
1066 n_elements in length. It is filled in with the pointers to the
1067 chunks.
1068
1069 In either case, independent_calloc returns this pointer array, or
1070 null if the allocation failed. If n_elements is zero and "chunks"
1071 is null, it returns a chunk representing an array with zero elements
1072 (which should be freed if not wanted).
1073
1074 Each element must be individually freed when it is no longer
1075 needed. If you'd like to instead be able to free all at once, you
1076 should instead use regular calloc and assign pointers into this
1077 space to represent elements. (In this case though, you cannot
1078 independently free elements.)
1079
1080 independent_calloc simplifies and speeds up implementations of many
1081 kinds of pools. It may also be useful when constructing large data
1082 structures that initially have a fixed number of fixed-sized nodes,
1083 but the number is not known at compile time, and some of the nodes
1084 may later need to be freed. For example:
1085
1086 struct Node { int item; struct Node* next; };
1087
1088 struct Node* build_list() {
1089 struct Node** pool;
1090 int n = read_number_of_nodes_needed();
1091 if (n <= 0) return 0;
1092 pool = (struct Node**)(independent_calloc(n, sizeof(struct Node), 0);
1093 if (pool == 0) die();
1094 // organize into a linked list...
1095 struct Node* first = pool[0];
1096 for (i = 0; i < n-1; ++i)
1097 pool[i]->next = pool[i+1];
1098 free(pool); // Can now free the array (or not, if it is needed later)
1099 return first;
1100 }
1101 */
1102 #if __STD_C
1103 Void_t** public_iCALLOc(size_t, size_t, Void_t**);
1104 #else
1105 Void_t** public_iCALLOc();
1106 #endif
1107
1108 /*
1109 independent_comalloc(size_t n_elements, size_t sizes[], Void_t* chunks[]);
1110
1111 independent_comalloc allocates, all at once, a set of n_elements
1112 chunks with sizes indicated in the "sizes" array. It returns
1113 an array of pointers to these elements, each of which can be
1114 independently freed, realloc'ed etc. The elements are guaranteed to
1115 be adjacently allocated (this is not guaranteed to occur with
1116 multiple callocs or mallocs), which may also improve cache locality
1117 in some applications.
1118
1119 The "chunks" argument is optional (i.e., may be null). If it is null
1120 the returned array is itself dynamically allocated and should also
1121 be freed when it is no longer needed. Otherwise, the chunks array
1122 must be of at least n_elements in length. It is filled in with the
1123 pointers to the chunks.
1124
1125 In either case, independent_comalloc returns this pointer array, or
1126 null if the allocation failed. If n_elements is zero and chunks is
1127 null, it returns a chunk representing an array with zero elements
1128 (which should be freed if not wanted).
1129
1130 Each element must be individually freed when it is no longer
1131 needed. If you'd like to instead be able to free all at once, you
1132 should instead use a single regular malloc, and assign pointers at
1133 particular offsets in the aggregate space. (In this case though, you
1134 cannot independently free elements.)
1135
1136 independent_comallac differs from independent_calloc in that each
1137 element may have a different size, and also that it does not
1138 automatically clear elements.
1139
1140 independent_comalloc can be used to speed up allocation in cases
1141 where several structs or objects must always be allocated at the
1142 same time. For example:
1143
1144 struct Head { ... }
1145 struct Foot { ... }
1146
1147 void send_message(char* msg) {
1148 int msglen = strlen(msg);
1149 size_t sizes[3] = { sizeof(struct Head), msglen, sizeof(struct Foot) };
1150 void* chunks[3];
1151 if (independent_comalloc(3, sizes, chunks) == 0)
1152 die();
1153 struct Head* head = (struct Head*)(chunks[0]);
1154 char* body = (char*)(chunks[1]);
1155 struct Foot* foot = (struct Foot*)(chunks[2]);
1156 // ...
1157 }
1158
1159 In general though, independent_comalloc is worth using only for
1160 larger values of n_elements. For small values, you probably won't
1161 detect enough difference from series of malloc calls to bother.
1162
1163 Overuse of independent_comalloc can increase overall memory usage,
1164 since it cannot reuse existing noncontiguous small chunks that
1165 might be available for some of the elements.
1166 */
1167 #if __STD_C
1168 Void_t** public_iCOMALLOc(size_t, size_t*, Void_t**);
1169 #else
1170 Void_t** public_iCOMALLOc();
1171 #endif
1172
1173
1174 /*
1175 pvalloc(size_t n);
1176 Equivalent to valloc(minimum-page-that-holds(n)), that is,
1177 round up n to nearest pagesize.
1178 */
1179 #if __STD_C
1180 Void_t* public_pVALLOc(size_t);
1181 #else
1182 Void_t* public_pVALLOc();
1183 #endif
1184
1185 /*
1186 cfree(Void_t* p);
1187 Equivalent to free(p).
1188
1189 cfree is needed/defined on some systems that pair it with calloc,
1190 for odd historical reasons (such as: cfree is used in example
1191 code in the first edition of K&R).
1192 */
1193 #if __STD_C
1194 void public_cFREe(Void_t*);
1195 #else
1196 void public_cFREe();
1197 #endif
1198
1199 /*
1200 malloc_trim(size_t pad);
1201
1202 If possible, gives memory back to the system (via negative
1203 arguments to sbrk) if there is unused memory at the `high' end of
1204 the malloc pool. You can call this after freeing large blocks of
1205 memory to potentially reduce the system-level memory requirements
1206 of a program. However, it cannot guarantee to reduce memory. Under
1207 some allocation patterns, some large free blocks of memory will be
1208 locked between two used chunks, so they cannot be given back to
1209 the system.
1210
1211 The `pad' argument to malloc_trim represents the amount of free
1212 trailing space to leave untrimmed. If this argument is zero,
1213 only the minimum amount of memory to maintain internal data
1214 structures will be left (one page or less). Non-zero arguments
1215 can be supplied to maintain enough trailing space to service
1216 future expected allocations without having to re-obtain memory
1217 from the system.
1218
1219 Malloc_trim returns 1 if it actually released any memory, else 0.
1220 On systems that do not support "negative sbrks", it will always
1221 rreturn 0.
1222 */
1223 #if __STD_C
1224 int public_mTRIm(size_t);
1225 #else
1226 int public_mTRIm();
1227 #endif
1228
1229 /*
1230 malloc_usable_size(Void_t* p);
1231
1232 Returns the number of bytes you can actually use in
1233 an allocated chunk, which may be more than you requested (although
1234 often not) due to alignment and minimum size constraints.
1235 You can use this many bytes without worrying about
1236 overwriting other allocated objects. This is not a particularly great
1237 programming practice. malloc_usable_size can be more useful in
1238 debugging and assertions, for example:
1239
1240 p = malloc(n);
1241 assert(malloc_usable_size(p) >= 256);
1242
1243 */
1244 #if __STD_C
1245 size_t public_mUSABLe(Void_t*);
1246 #else
1247 size_t public_mUSABLe();
1248 #endif
1249
1250 /*
1251 malloc_stats();
1252 Prints on stderr the amount of space obtained from the system (both
1253 via sbrk and mmap), the maximum amount (which may be more than
1254 current if malloc_trim and/or munmap got called), and the current
1255 number of bytes allocated via malloc (or realloc, etc) but not yet
1256 freed. Note that this is the number of bytes allocated, not the
1257 number requested. It will be larger than the number requested
1258 because of alignment and bookkeeping overhead. Because it includes
1259 alignment wastage as being in use, this figure may be greater than
1260 zero even when no user-level chunks are allocated.
1261
1262 The reported current and maximum system memory can be inaccurate if
1263 a program makes other calls to system memory allocation functions
1264 (normally sbrk) outside of malloc.
1265
1266 malloc_stats prints only the most commonly interesting statistics.
1267 More information can be obtained by calling mallinfo.
1268
1269 */
1270 #if __STD_C
1271 void public_mSTATs();
1272 #else
1273 void public_mSTATs();
1274 #endif
1275
1276 /* mallopt tuning options */
1277
1278 /*
1279 M_MXFAST is the maximum request size used for "fastbins", special bins
1280 that hold returned chunks without consolidating their spaces. This
1281 enables future requests for chunks of the same size to be handled
1282 very quickly, but can increase fragmentation, and thus increase the
1283 overall memory footprint of a program.
1284
1285 This malloc manages fastbins very conservatively yet still
1286 efficiently, so fragmentation is rarely a problem for values less
1287 than or equal to the default. The maximum supported value of MXFAST
1288 is 80. You wouldn't want it any higher than this anyway. Fastbins
1289 are designed especially for use with many small structs, objects or
1290 strings -- the default handles structs/objects/arrays with sizes up
1291 to 16 4byte fields, or small strings representing words, tokens,
1292 etc. Using fastbins for larger objects normally worsens
1293 fragmentation without improving speed.
1294
1295 M_MXFAST is set in REQUEST size units. It is internally used in
1296 chunksize units, which adds padding and alignment. You can reduce
1297 M_MXFAST to 0 to disable all use of fastbins. This causes the malloc
1298 algorithm to be a closer approximation of fifo-best-fit in all cases,
1299 not just for larger requests, but will generally cause it to be
1300 slower.
1301 */
1302
1303
1304 /* M_MXFAST is a standard SVID/XPG tuning option, usually listed in malloc.h */
1305 #ifndef M_MXFAST
1306 #define M_MXFAST 1
1307 #endif
1308
1309 #ifndef DEFAULT_MXFAST
1310 #define DEFAULT_MXFAST 64
1311 #endif
1312
1313
1314 /*
1315 M_TRIM_THRESHOLD is the maximum amount of unused top-most memory
1316 to keep before releasing via malloc_trim in free().
1317
1318 Automatic trimming is mainly useful in long-lived programs.
1319 Because trimming via sbrk can be slow on some systems, and can
1320 sometimes be wasteful (in cases where programs immediately
1321 afterward allocate more large chunks) the value should be high
1322 enough so that your overall system performance would improve by
1323 releasing this much memory.
1324
1325 The trim threshold and the mmap control parameters (see below)
1326 can be traded off with one another. Trimming and mmapping are
1327 two different ways of releasing unused memory back to the
1328 system. Between these two, it is often possible to keep
1329 system-level demands of a long-lived program down to a bare
1330 minimum. For example, in one test suite of sessions measuring
1331 the XF86 X server on Linux, using a trim threshold of 128K and a
1332 mmap threshold of 192K led to near-minimal long term resource
1333 consumption.
1334
1335 If you are using this malloc in a long-lived program, it should
1336 pay to experiment with these values. As a rough guide, you
1337 might set to a value close to the average size of a process
1338 (program) running on your system. Releasing this much memory
1339 would allow such a process to run in memory. Generally, it's
1340 worth it to tune for trimming rather tham memory mapping when a
1341 program undergoes phases where several large chunks are
1342 allocated and released in ways that can reuse each other's
1343 storage, perhaps mixed with phases where there are no such
1344 chunks at all. And in well-behaved long-lived programs,
1345 controlling release of large blocks via trimming versus mapping
1346 is usually faster.
1347
1348 However, in most programs, these parameters serve mainly as
1349 protection against the system-level effects of carrying around
1350 massive amounts of unneeded memory. Since frequent calls to
1351 sbrk, mmap, and munmap otherwise degrade performance, the default
1352 parameters are set to relatively high values that serve only as
1353 safeguards.
1354
1355 The trim value must be greater than page size to have any useful
1356 effect. To disable trimming completely, you can set to
1357 (unsigned long)(-1)
1358
1359 Trim settings interact with fastbin (MXFAST) settings: Unless
1360 TRIM_FASTBINS is defined, automatic trimming never takes place upon
1361 freeing a chunk with size less than or equal to MXFAST. Trimming is
1362 instead delayed until subsequent freeing of larger chunks. However,
1363 you can still force an attempted trim by calling malloc_trim.
1364
1365 Also, trimming is not generally possible in cases where
1366 the main arena is obtained via mmap.
1367
1368 Note that the trick some people use of mallocing a huge space and
1369 then freeing it at program startup, in an attempt to reserve system
1370 memory, doesn't have the intended effect under automatic trimming,
1371 since that memory will immediately be returned to the system.
1372 */
1373
1374 #define M_TRIM_THRESHOLD -1
1375
1376 #ifndef DEFAULT_TRIM_THRESHOLD
1377 #define DEFAULT_TRIM_THRESHOLD (256 * 1024)
1378 #endif
1379
1380 /*
1381 M_TOP_PAD is the amount of extra `padding' space to allocate or
1382 retain whenever sbrk is called. It is used in two ways internally:
1383
1384 * When sbrk is called to extend the top of the arena to satisfy
1385 a new malloc request, this much padding is added to the sbrk
1386 request.
1387
1388 * When malloc_trim is called automatically from free(),
1389 it is used as the `pad' argument.
1390
1391 In both cases, the actual amount of padding is rounded
1392 so that the end of the arena is always a system page boundary.
1393
1394 The main reason for using padding is to avoid calling sbrk so
1395 often. Having even a small pad greatly reduces the likelihood
1396 that nearly every malloc request during program start-up (or
1397 after trimming) will invoke sbrk, which needlessly wastes
1398 time.
1399
1400 Automatic rounding-up to page-size units is normally sufficient
1401 to avoid measurable overhead, so the default is 0. However, in
1402 systems where sbrk is relatively slow, it can pay to increase
1403 this value, at the expense of carrying around more memory than
1404 the program needs.
1405 */
1406
1407 #define M_TOP_PAD -2
1408
1409 #ifndef DEFAULT_TOP_PAD
1410 #define DEFAULT_TOP_PAD (0)
1411 #endif
1412
1413 /*
1414 M_MMAP_THRESHOLD is the request size threshold for using mmap()
1415 to service a request. Requests of at least this size that cannot
1416 be allocated using already-existing space will be serviced via mmap.
1417 (If enough normal freed space already exists it is used instead.)
1418
1419 Using mmap segregates relatively large chunks of memory so that
1420 they can be individually obtained and released from the host
1421 system. A request serviced through mmap is never reused by any
1422 other request (at least not directly; the system may just so
1423 happen to remap successive requests to the same locations).
1424
1425 Segregating space in this way has the benefits that:
1426
1427 1. Mmapped space can ALWAYS be individually released back
1428 to the system, which helps keep the system level memory
1429 demands of a long-lived program low.
1430 2. Mapped memory can never become `locked' between
1431 other chunks, as can happen with normally allocated chunks, which
1432 means that even trimming via malloc_trim would not release them.
1433 3. On some systems with "holes" in address spaces, mmap can obtain
1434 memory that sbrk cannot.
1435
1436 However, it has the disadvantages that:
1437
1438 1. The space cannot be reclaimed, consolidated, and then
1439 used to service later requests, as happens with normal chunks.
1440 2. It can lead to more wastage because of mmap page alignment
1441 requirements
1442 3. It causes malloc performance to be more dependent on host
1443 system memory management support routines which may vary in
1444 implementation quality and may impose arbitrary
1445 limitations. Generally, servicing a request via normal
1446 malloc steps is faster than going through a system's mmap.
1447
1448 The advantages of mmap nearly always outweigh disadvantages for
1449 "large" chunks, but the value of "large" varies across systems. The
1450 default is an empirically derived value that works well in most
1451 systems.
1452 */
1453
1454 #define M_MMAP_THRESHOLD -3
1455
1456 #ifndef DEFAULT_MMAP_THRESHOLD
1457 #define DEFAULT_MMAP_THRESHOLD (256 * 1024)
1458 #endif
1459
1460 /*
1461 M_MMAP_MAX is the maximum number of requests to simultaneously
1462 service using mmap. This parameter exists because
1463 . Some systems have a limited number of internal tables for
1464 use by mmap, and using more than a few of them may degrade
1465 performance.
1466
1467 The default is set to a value that serves only as a safeguard.
1468 Setting to 0 disables use of mmap for servicing large requests. If
1469 HAVE_MMAP is not set, the default value is 0, and attempts to set it
1470 to non-zero values in mallopt will fail.
1471 */
1472
1473 #define M_MMAP_MAX -4
1474
1475 #ifndef DEFAULT_MMAP_MAX
1476 #if HAVE_MMAP
1477 #define DEFAULT_MMAP_MAX (65536)
1478 #else
1479 #define DEFAULT_MMAP_MAX (0)
1480 #endif
1481 #endif
1482
1483 #ifdef __cplusplus
1484 }; /* end of extern "C" */
1485 #endif
1486
1487 /*
1488 ========================================================================
1489 To make a fully customizable malloc.h header file, cut everything
1490 above this line, put into file malloc.h, edit to suit, and #include it
1491 on the next line, as well as in programs that use this malloc.
1492 ========================================================================
1493 */
1494
1495 /* #include "malloc.h" */
1496
1497 /* --------------------- public wrappers ---------------------- */
1498
1499 #ifdef USE_PUBLIC_MALLOC_WRAPPERS
1500
1501 /* Declare all routines as internal */
1502 #if __STD_C
1503 static Void_t* mALLOc(size_t);
1504 static void fREe(Void_t*);
1505 static Void_t* rEALLOc(Void_t*, size_t);
1506 static Void_t* mEMALIGn(size_t, size_t);
1507 static Void_t* vALLOc(size_t);
1508 static Void_t* pVALLOc(size_t);
1509 static Void_t* cALLOc(size_t, size_t);
1510 static Void_t** iCALLOc(size_t, size_t, Void_t**);
1511 static Void_t** iCOMALLOc(size_t, size_t*, Void_t**);
1512 static void cFREe(Void_t*);
1513 static int mTRIm(size_t);
1514 static size_t mUSABLe(Void_t*);
1515 static void mSTATs();
1516 static int mALLOPt(int, int);
1517 static struct mallinfo mALLINFo(void);
1518 #else
1519 static Void_t* mALLOc();
1520 static void fREe();
1521 static Void_t* rEALLOc();
1522 static Void_t* mEMALIGn();
1523 static Void_t* vALLOc();
1524 static Void_t* pVALLOc();
1525 static Void_t* cALLOc();
1526 static Void_t** iCALLOc();
1527 static Void_t** iCOMALLOc();
1528 static void cFREe();
1529 static int mTRIm();
1530 static size_t mUSABLe();
1531 static void mSTATs();
1532 static int mALLOPt();
1533 static struct mallinfo mALLINFo();
1534 #endif
1535
1536 /*
1537 MALLOC_PREACTION and MALLOC_POSTACTION should be
1538 defined to return 0 on success, and nonzero on failure.
1539 The return value of MALLOC_POSTACTION is currently ignored
1540 in wrapper functions since there is no reasonable default
1541 action to take on failure.
1542 */
1543
1544
1545 #ifdef USE_MALLOC_LOCK
1546
1547 #ifdef WIN32
1548
1549 static int mALLOC_MUTEx;
1550 #define MALLOC_PREACTION slwait(&mALLOC_MUTEx)
1551 #define MALLOC_POSTACTION slrelease(&mALLOC_MUTEx)
1552
1553 #else
1554
1555 #include <pthread.h>
1556
1557 static pthread_mutex_t mALLOC_MUTEx = PTHREAD_MUTEX_INITIALIZER;
1558
1559 #define MALLOC_PREACTION pthread_mutex_lock(&mALLOC_MUTEx)
1560 #define MALLOC_POSTACTION pthread_mutex_unlock(&mALLOC_MUTEx)
1561
1562 #endif /* USE_MALLOC_LOCK */
1563
1564 #else
1565
1566 /* Substitute anything you like for these */
1567
1568 #define MALLOC_PREACTION (0)
1569 #define MALLOC_POSTACTION (0)
1570
1571 #endif
1572
1573 Void_t* public_mALLOc(size_t bytes) {
1574 Void_t* m;
1575 if (MALLOC_PREACTION != 0) {
1576 return 0;
1577 }
1578 m = mALLOc(bytes);
1579 if (MALLOC_POSTACTION != 0) {
1580 }
1581 return m;
1582 }
1583
1584 void public_fREe(Void_t* m) {
1585 if (MALLOC_PREACTION != 0) {
1586 return;
1587 }
1588 fREe(m);
1589 if (MALLOC_POSTACTION != 0) {
1590 }
1591 }
1592
1593 Void_t* public_rEALLOc(Void_t* m, size_t bytes) {
1594 if (MALLOC_PREACTION != 0) {
1595 return 0;
1596 }
1597 m = rEALLOc(m, bytes);
1598 if (MALLOC_POSTACTION != 0) {
1599 }
1600 return m;
1601 }
1602
1603 Void_t* public_mEMALIGn(size_t alignment, size_t bytes) {
1604 Void_t* m;
1605 if (MALLOC_PREACTION != 0) {
1606 return 0;
1607 }
1608 m = mEMALIGn(alignment, bytes);
1609 if (MALLOC_POSTACTION != 0) {
1610 }
1611 return m;
1612 }
1613
1614 Void_t* public_vALLOc(size_t bytes) {
1615 Void_t* m;
1616 if (MALLOC_PREACTION != 0) {
1617 return 0;
1618 }
1619 m = vALLOc(bytes);
1620 if (MALLOC_POSTACTION != 0) {
1621 }
1622 return m;
1623 }
1624
1625 Void_t* public_pVALLOc(size_t bytes) {
1626 Void_t* m;
1627 if (MALLOC_PREACTION != 0) {
1628 return 0;
1629 }
1630 m = pVALLOc(bytes);
1631 if (MALLOC_POSTACTION != 0) {
1632 }
1633 return m;
1634 }
1635
1636 Void_t* public_cALLOc(size_t n, size_t elem_size) {
1637 Void_t* m;
1638 if (MALLOC_PREACTION != 0) {
1639 return 0;
1640 }
1641 m = cALLOc(n, elem_size);
1642 if (MALLOC_POSTACTION != 0) {
1643 }
1644 return m;
1645 }
1646
1647
1648 Void_t** public_iCALLOc(size_t n, size_t elem_size, Void_t** chunks) {
1649 Void_t** m;
1650 if (MALLOC_PREACTION != 0) {
1651 return 0;
1652 }
1653 m = iCALLOc(n, elem_size, chunks);
1654 if (MALLOC_POSTACTION != 0) {
1655 }
1656 return m;
1657 }
1658
1659 Void_t** public_iCOMALLOc(size_t n, size_t sizes[], Void_t** chunks) {
1660 Void_t** m;
1661 if (MALLOC_PREACTION != 0) {
1662 return 0;
1663 }
1664 m = iCOMALLOc(n, sizes, chunks);
1665 if (MALLOC_POSTACTION != 0) {
1666 }
1667 return m;
1668 }
1669
1670 void public_cFREe(Void_t* m) {
1671 if (MALLOC_PREACTION != 0) {
1672 return;
1673 }
1674 cFREe(m);
1675 if (MALLOC_POSTACTION != 0) {
1676 }
1677 }
1678
1679 int public_mTRIm(size_t s) {
1680 int result;
1681 if (MALLOC_PREACTION != 0) {
1682 return 0;
1683 }
1684 result = mTRIm(s);
1685 if (MALLOC_POSTACTION != 0) {
1686 }
1687 return result;
1688 }
1689
1690 size_t public_mUSABLe(Void_t* m) {
1691 size_t result;
1692 if (MALLOC_PREACTION != 0) {
1693 return 0;
1694 }
1695 result = mUSABLe(m);
1696 if (MALLOC_POSTACTION != 0) {
1697 }
1698 return result;
1699 }
1700
1701 void public_mSTATs() {
1702 if (MALLOC_PREACTION != 0) {
1703 return;
1704 }
1705 mSTATs();
1706 if (MALLOC_POSTACTION != 0) {
1707 }
1708 }
1709
1710 struct mallinfo public_mALLINFo() {
1711 struct mallinfo m;
1712 if (MALLOC_PREACTION != 0) {
1713 struct mallinfo nm = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
1714 return nm;
1715 }
1716 m = mALLINFo();
1717 if (MALLOC_POSTACTION != 0) {
1718 }
1719 return m;
1720 }
1721
1722 int public_mALLOPt(int p, int v) {
1723 int result;
1724 if (MALLOC_PREACTION != 0) {
1725 return 0;
1726 }
1727 result = mALLOPt(p, v);
1728 if (MALLOC_POSTACTION != 0) {
1729 }
1730 return result;
1731 }
1732
1733 #endif
1734
1735
1736
1737 /* ------------- Optional versions of memcopy ---------------- */
1738
1739
1740 #if USE_MEMCPY
1741
1742 /*
1743 Note: memcpy is ONLY invoked with non-overlapping regions,
1744 so the (usually slower) memmove is not needed.
1745 */
1746
1747 #define MALLOC_COPY(dest, src, nbytes) memcpy(dest, src, nbytes)
1748 #define MALLOC_ZERO(dest, nbytes) memset(dest, 0, nbytes)
1749
1750 #else /* !USE_MEMCPY */
1751
1752 /* Use Duff's device for good zeroing/copying performance. */
1753
1754 #define MALLOC_ZERO(charp, nbytes) \
1755 do { \
1756 INTERNAL_SIZE_T* mzp = (INTERNAL_SIZE_T*)(charp); \
1757 CHUNK_SIZE_T mctmp = (nbytes)/sizeof(INTERNAL_SIZE_T); \
1758 long mcn; \
1759 if (mctmp < 8) mcn = 0; else { mcn = (mctmp-1)/8; mctmp %= 8; } \
1760 switch (mctmp) { \
1761 case 0: for(;;) { *mzp++ = 0; \
1762 case 7: *mzp++ = 0; \
1763 case 6: *mzp++ = 0; \
1764 case 5: *mzp++ = 0; \
1765 case 4: *mzp++ = 0; \
1766 case 3: *mzp++ = 0; \
1767 case 2: *mzp++ = 0; \
1768 case 1: *mzp++ = 0; if(mcn <= 0) break; mcn--; } \
1769 } \
1770 } while(0)
1771
1772 #define MALLOC_COPY(dest,src,nbytes) \
1773 do { \
1774 INTERNAL_SIZE_T* mcsrc = (INTERNAL_SIZE_T*) src; \
1775 INTERNAL_SIZE_T* mcdst = (INTERNAL_SIZE_T*) dest; \
1776 CHUNK_SIZE_T mctmp = (nbytes)/sizeof(INTERNAL_SIZE_T); \
1777 long mcn; \
1778 if (mctmp < 8) mcn = 0; else { mcn = (mctmp-1)/8; mctmp %= 8; } \
1779 switch (mctmp) { \
1780 case 0: for(;;) { *mcdst++ = *mcsrc++; \
1781 case 7: *mcdst++ = *mcsrc++; \
1782 case 6: *mcdst++ = *mcsrc++; \
1783 case 5: *mcdst++ = *mcsrc++; \
1784 case 4: *mcdst++ = *mcsrc++; \
1785 case 3: *mcdst++ = *mcsrc++; \
1786 case 2: *mcdst++ = *mcsrc++; \
1787 case 1: *mcdst++ = *mcsrc++; if(mcn <= 0) break; mcn--; } \
1788 } \
1789 } while(0)
1790
1791 #endif
1792
1793 /* ------------------ MMAP support ------------------ */
1794
1795
1796 #if HAVE_MMAP
1797
1798 #ifndef LACKS_FCNTL_H
1799 #include <fcntl.h>
1800 #endif
1801
1802 #ifndef LACKS_SYS_MMAN_H
1803 #include <sys/mman.h>
1804 #endif
1805
1806 #if !defined(MAP_ANONYMOUS) && defined(MAP_ANON)
1807 #define MAP_ANONYMOUS MAP_ANON
1808 #endif
1809
1810 /*
1811 Nearly all versions of mmap support MAP_ANONYMOUS,
1812 so the following is unlikely to be needed, but is
1813 supplied just in case.
1814 */
1815
1816 #ifndef MAP_ANONYMOUS
1817
1818 static int dev_zero_fd = -1; /* Cached file descriptor for /dev/zero. */
1819
1820 #define MMAP(addr, size, prot, flags) ((dev_zero_fd < 0) ? \
1821 (dev_zero_fd = open("/dev/zero", O_RDWR), \
1822 mmap((addr), (size), (prot), (flags), dev_zero_fd, 0)) : \
1823 mmap((addr), (size), (prot), (flags), dev_zero_fd, 0))
1824
1825 #else
1826
1827 #define MMAP(addr, size, prot, flags) \
1828 (mmap((addr), (size), (prot), (flags)|MAP_ANONYMOUS, -1, 0))
1829
1830 #endif
1831
1832
1833 #endif /* HAVE_MMAP */
1834
1835
1836 /*
1837 ----------------------- Chunk representations -----------------------
1838 */
1839
1840
1841 /*
1842 This struct declaration is misleading (but accurate and necessary).
1843 It declares a "view" into memory allowing access to necessary
1844 fields at known offsets from a given base. See explanation below.
1845 */
1846
1847 struct malloc_chunk {
1848
1849 INTERNAL_SIZE_T prev_size; /* Size of previous chunk (if free). */
1850 INTERNAL_SIZE_T size; /* Size in bytes, including overhead. */
1851
1852 struct malloc_chunk* fd; /* double links -- used only if free. */
1853 struct malloc_chunk* bk;
1854 };
1855
1856
1857 typedef struct malloc_chunk* mchunkptr;
1858
1859 /*
1860 malloc_chunk details:
1861
1862 (The following includes lightly edited explanations by Colin Plumb.)
1863
1864 Chunks of memory are maintained using a `boundary tag' method as
1865 described in e.g., Knuth or Standish. (See the paper by Paul
1866 Wilson ftp://ftp.cs.utexas.edu/pub/garbage/allocsrv.ps for a
1867 survey of such techniques.) Sizes of free chunks are stored both
1868 in the front of each chunk and at the end. This makes
1869 consolidating fragmented chunks into bigger chunks very fast. The
1870 size fields also hold bits representing whether chunks are free or
1871 in use.
1872
1873 An allocated chunk looks like this:
1874
1875
1876 chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1877 | Size of previous chunk, if allocated | |
1878 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1879 | Size of chunk, in bytes |P|
1880 mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1881 | User data starts here... .
1882 . .
1883 . (malloc_usable_space() bytes) .
1884 . |
1885 nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1886 | Size of chunk |
1887 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1888
1889
1890 Where "chunk" is the front of the chunk for the purpose of most of
1891 the malloc code, but "mem" is the pointer that is returned to the
1892 user. "Nextchunk" is the beginning of the next contiguous chunk.
1893
1894 Chunks always begin on even word boundries, so the mem portion
1895 (which is returned to the user) is also on an even word boundary, and
1896 thus at least double-word aligned.
1897
1898 Free chunks are stored in circular doubly-linked lists, and look like this:
1899
1900 chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1901 | Size of previous chunk |
1902 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1903 `head:' | Size of chunk, in bytes |P|
1904 mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1905 | Forward pointer to next chunk in list |
1906 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1907 | Back pointer to previous chunk in list |
1908 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1909 | Unused space (may be 0 bytes long) .
1910 . .
1911 . |
1912 nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1913 `foot:' | Size of chunk, in bytes |
1914 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1915
1916 The P (PREV_INUSE) bit, stored in the unused low-order bit of the
1917 chunk size (which is always a multiple of two words), is an in-use
1918 bit for the *previous* chunk. If that bit is *clear*, then the
1919 word before the current chunk size contains the previous chunk
1920 size, and can be used to find the front of the previous chunk.
1921 The very first chunk allocated always has this bit set,
1922 preventing access to non-existent (or non-owned) memory. If
1923 prev_inuse is set for any given chunk, then you CANNOT determine
1924 the size of the previous chunk, and might even get a memory
1925 addressing fault when trying to do so.
1926
1927 Note that the `foot' of the current chunk is actually represented
1928 as the prev_size of the NEXT chunk. This makes it easier to
1929 deal with alignments etc but can be very confusing when trying
1930 to extend or adapt this code.
1931
1932 The two exceptions to all this are
1933
1934 1. The special chunk `top' doesn't bother using the
1935 trailing size field since there is no next contiguous chunk
1936 that would have to index off it. After initialization, `top'
1937 is forced to always exist. If it would become less than
1938 MINSIZE bytes long, it is replenished.
1939
1940 2. Chunks allocated via mmap, which have the second-lowest-order
1941 bit (IS_MMAPPED) set in their size fields. Because they are
1942 allocated one-by-one, each must contain its own trailing size field.
1943
1944 */
1945
1946 /*
1947 ---------- Size and alignment checks and conversions ----------
1948 */
1949
1950 /* conversion from malloc headers to user pointers, and back */
1951
1952 #define chunk2mem(p) ((Void_t*)((char*)(p) + 2*SIZE_SZ))
1953 #define mem2chunk(mem) ((mchunkptr)((char*)(mem) - 2*SIZE_SZ))
1954
1955 /* The smallest possible chunk */
1956 #define MIN_CHUNK_SIZE (sizeof(struct malloc_chunk))
1957
1958 /* The smallest size we can malloc is an aligned minimal chunk */
1959
1960 #define MINSIZE \
1961 (CHUNK_SIZE_T)(((MIN_CHUNK_SIZE+MALLOC_ALIGN_MASK) & ~MALLOC_ALIGN_MASK))
1962
1963 /* Check if m has acceptable alignment */
1964
1965 #define aligned_OK(m) (((PTR_UINT)((m)) & (MALLOC_ALIGN_MASK)) == 0)
1966
1967
1968 /*
1969 Check if a request is so large that it would wrap around zero when
1970 padded and aligned. To simplify some other code, the bound is made
1971 low enough so that adding MINSIZE will also not wrap around sero.
1972 */
1973
1974 #define REQUEST_OUT_OF_RANGE(req) \
1975 ((CHUNK_SIZE_T)(req) >= \
1976 (CHUNK_SIZE_T)(INTERNAL_SIZE_T)(-2 * MINSIZE))
1977
1978 /* pad request bytes into a usable size -- internal version */
1979
1980 #define request2size(req) \
1981 (((req) + SIZE_SZ + MALLOC_ALIGN_MASK < MINSIZE) ? \
1982 MINSIZE : \
1983 ((req) + SIZE_SZ + MALLOC_ALIGN_MASK) & ~MALLOC_ALIGN_MASK)
1984
1985 /* Same, except also perform argument check */
1986
1987 #define checked_request2size(req, sz) \
1988 if (REQUEST_OUT_OF_RANGE(req)) { \
1989 MALLOC_FAILURE_ACTION; \
1990 return 0; \
1991 } \
1992 (sz) = request2size(req);
1993
1994 /*
1995 --------------- Physical chunk operations ---------------
1996 */
1997
1998
1999 /* size field is or'ed with PREV_INUSE when previous adjacent chunk in use */
2000 #define PREV_INUSE 0x1
2001
2002 /* extract inuse bit of previous chunk */
2003 #define prev_inuse(p) ((p)->size & PREV_INUSE)
2004
2005
2006 /* size field is or'ed with IS_MMAPPED if the chunk was obtained with mmap() */
2007 #define IS_MMAPPED 0x2
2008
2009 /* check for mmap()'ed chunk */
2010 #define chunk_is_mmapped(p) ((p)->size & IS_MMAPPED)
2011
2012 /*
2013 Bits to mask off when extracting size
2014
2015 Note: IS_MMAPPED is intentionally not masked off from size field in
2016 macros for which mmapped chunks should never be seen. This should
2017 cause helpful core dumps to occur if it is tried by accident by
2018 people extending or adapting this malloc.
2019 */
2020 #define SIZE_BITS (PREV_INUSE|IS_MMAPPED)
2021
2022 /* Get size, ignoring use bits */
2023 #define chunksize(p) ((p)->size & ~(SIZE_BITS))
2024
2025
2026 /* Ptr to next physical malloc_chunk. */
2027 #define next_chunk(p) ((mchunkptr)( ((char*)(p)) + ((p)->size & ~PREV_INUSE) ))
2028
2029 /* Ptr to previous physical malloc_chunk */
2030 #define prev_chunk(p) ((mchunkptr)( ((char*)(p)) - ((p)->prev_size) ))
2031
2032 /* Treat space at ptr + offset as a chunk */
2033 #define chunk_at_offset(p, s) ((mchunkptr)(((char*)(p)) + (s)))
2034
2035 /* extract p's inuse bit */
2036 #define inuse(p)\
2037 ((((mchunkptr)(((char*)(p))+((p)->size & ~PREV_INUSE)))->size) & PREV_INUSE)
2038
2039 /* set/clear chunk as being inuse without otherwise disturbing */
2040 #define set_inuse(p)\
2041 ((mchunkptr)(((char*)(p)) + ((p)->size & ~PREV_INUSE)))->size |= PREV_INUSE
2042
2043 #define clear_inuse(p)\
2044 ((mchunkptr)(((char*)(p)) + ((p)->size & ~PREV_INUSE)))->size &= ~(PREV_INUSE)
2045
2046
2047 /* check/set/clear inuse bits in known places */
2048 #define inuse_bit_at_offset(p, s)\
2049 (((mchunkptr)(((char*)(p)) + (s)))->size & PREV_INUSE)
2050
2051 #define set_inuse_bit_at_offset(p, s)\
2052 (((mchunkptr)(((char*)(p)) + (s)))->size |= PREV_INUSE)
2053
2054 #define clear_inuse_bit_at_offset(p, s)\
2055 (((mchunkptr)(((char*)(p)) + (s)))->size &= ~(PREV_INUSE))
2056
2057
2058 /* Set size at head, without disturbing its use bit */
2059 #define set_head_size(p, s) ((p)->size = (((p)->size & PREV_INUSE) | (s)))
2060
2061 /* Set size/use field */
2062 #define set_head(p, s) ((p)->size = (s))
2063
2064 /* Set size at footer (only when chunk is not in use) */
2065 #define set_foot(p, s) (((mchunkptr)((char*)(p) + (s)))->prev_size = (s))
2066
2067
2068 /*
2069 -------------------- Internal data structures --------------------
2070
2071 All internal state is held in an instance of malloc_state defined
2072 below. There are no other static variables, except in two optional
2073 cases:
2074 * If USE_MALLOC_LOCK is defined, the mALLOC_MUTEx declared above.
2075 * If HAVE_MMAP is true, but mmap doesn't support
2076 MAP_ANONYMOUS, a dummy file descriptor for mmap.
2077
2078 Beware of lots of tricks that minimize the total bookkeeping space
2079 requirements. The result is a little over 1K bytes (for 4byte
2080 pointers and size_t.)
2081 */
2082
2083 /*
2084 Bins
2085
2086 An array of bin headers for free chunks. Each bin is doubly
2087 linked. The bins are approximately proportionally (log) spaced.
2088 There are a lot of these bins (128). This may look excessive, but
2089 works very well in practice. Most bins hold sizes that are
2090 unusual as malloc request sizes, but are more usual for fragments
2091 and consolidated sets of chunks, which is what these bins hold, so
2092 they can be found quickly. All procedures maintain the invariant
2093 that no consolidated chunk physically borders another one, so each
2094 chunk in a list is known to be preceeded and followed by either
2095 inuse chunks or the ends of memory.
2096
2097 Chunks in bins are kept in size order, with ties going to the
2098 approximately least recently used chunk. Ordering isn't needed
2099 for the small bins, which all contain the same-sized chunks, but
2100 facilitates best-fit allocation for larger chunks. These lists
2101 are just sequential. Keeping them in order almost never requires
2102 enough traversal to warrant using fancier ordered data
2103 structures.
2104
2105 Chunks of the same size are linked with the most
2106 recently freed at the front, and allocations are taken from the
2107 back. This results in LRU (FIFO) allocation order, which tends
2108 to give each chunk an equal opportunity to be consolidated with
2109 adjacent freed chunks, resulting in larger free chunks and less
2110 fragmentation.
2111
2112 To simplify use in double-linked lists, each bin header acts
2113 as a malloc_chunk. This avoids special-casing for headers.
2114 But to conserve space and improve locality, we allocate
2115 only the fd/bk pointers of bins, and then use repositioning tricks
2116 to treat these as the fields of a malloc_chunk*.
2117 */
2118
2119 typedef struct malloc_chunk* mbinptr;
2120
2121 /* addressing -- note that bin_at(0) does not exist */
2122 #define bin_at(m, i) ((mbinptr)((char*)&((m)->bins[(i)<<1]) - (SIZE_SZ<<1)))
2123
2124 /* analog of ++bin */
2125 #define next_bin(b) ((mbinptr)((char*)(b) + (sizeof(mchunkptr)<<1)))
2126
2127 /* Reminders about list directionality within bins */
2128 #define first(b) ((b)->fd)
2129 #define last(b) ((b)->bk)
2130
2131 /* Take a chunk off a bin list */
2132 #define unlink(P, BK, FD) { \
2133 FD = P->fd; \
2134 BK = P->bk; \
2135 FD->bk = BK; \
2136 BK->fd = FD; \
2137 }
2138
2139 /*
2140 Indexing
2141
2142 Bins for sizes < 512 bytes contain chunks of all the same size, spaced
2143 8 bytes apart. Larger bins are approximately logarithmically spaced:
2144
2145 64 bins of size 8
2146 32 bins of size 64
2147 16 bins of size 512
2148 8 bins of size 4096
2149 4 bins of size 32768
2150 2 bins of size 262144
2151 1 bin of size what's left
2152
2153 The bins top out around 1MB because we expect to service large
2154 requests via mmap.
2155 */
2156
2157 #define NBINS 96
2158 #define NSMALLBINS 32
2159 #define SMALLBIN_WIDTH 8
2160 #define MIN_LARGE_SIZE 256
2161
2162 #define in_smallbin_range(sz) \
2163 ((CHUNK_SIZE_T)(sz) < (CHUNK_SIZE_T)MIN_LARGE_SIZE)
2164
2165 #define smallbin_index(sz) (((unsigned)(sz)) >> 3)
2166
2167 /*
2168 Compute index for size. We expect this to be inlined when
2169 compiled with optimization, else not, which works out well.
2170 */
2171 static int largebin_index(unsigned int sz) {
2172 unsigned int x = sz >> SMALLBIN_WIDTH;
2173 unsigned int m; /* bit position of highest set bit of m */
2174
2175 if (x >= 0x10000) return NBINS-1;
2176
2177 /* On intel, use BSRL instruction to find highest bit */
2178 #if defined(__GNUC__) && defined(i386)
2179
2180 __asm__("bsrl %1,%0\n\t"
2181 : "=r" (m)
2182 : "g" (x));
2183
2184 #else
2185 {
2186 /*
2187 Based on branch-free nlz algorithm in chapter 5 of Henry
2188 S. Warren Jr's book "Hacker's Delight".
2189 */
2190
2191 unsigned int n = ((x - 0x100) >> 16) & 8;
2192 x <<= n;
2193 m = ((x - 0x1000) >> 16) & 4;
2194 n += m;
2195 x <<= m;
2196 m = ((x - 0x4000) >> 16) & 2;
2197 n += m;
2198 x = (x << m) >> 14;
2199 m = 13 - n + (x & ~(x>>1));
2200 }
2201 #endif
2202
2203 /* Use next 2 bits to create finer-granularity bins */
2204 return NSMALLBINS + (m << 2) + ((sz >> (m + 6)) & 3);
2205 }
2206
2207 #define bin_index(sz) \
2208 ((in_smallbin_range(sz)) ? smallbin_index(sz) : largebin_index(sz))
2209
2210 /*
2211 FIRST_SORTED_BIN_SIZE is the chunk size corresponding to the
2212 first bin that is maintained in sorted order. This must
2213 be the smallest size corresponding to a given bin.
2214
2215 Normally, this should be MIN_LARGE_SIZE. But you can weaken
2216 best fit guarantees to sometimes speed up malloc by increasing value.
2217 Doing this means that malloc may choose a chunk that is
2218 non-best-fitting by up to the width of the bin.
2219
2220 Some useful cutoff values:
2221 512 - all bins sorted
2222 2560 - leaves bins <= 64 bytes wide unsorted
2223 12288 - leaves bins <= 512 bytes wide unsorted
2224 65536 - leaves bins <= 4096 bytes wide unsorted
2225 262144 - leaves bins <= 32768 bytes wide unsorted
2226 -1 - no bins sorted (not recommended!)
2227 */
2228
2229 #define FIRST_SORTED_BIN_SIZE MIN_LARGE_SIZE
2230 /* #define FIRST_SORTED_BIN_SIZE 65536 */
2231
2232 /*
2233 Unsorted chunks
2234
2235 All remainders from chunk splits, as well as all returned chunks,
2236 are first placed in the "unsorted" bin. They are then placed
2237 in regular bins after malloc gives them ONE chance to be used before
2238 binning. So, basically, the unsorted_chunks list acts as a queue,
2239 with chunks being placed on it in free (and malloc_consolidate),
2240 and taken off (to be either used or placed in bins) in malloc.
2241 */
2242
2243 /* The otherwise unindexable 1-bin is used to hold unsorted chunks. */
2244 #define unsorted_chunks(M) (bin_at(M, 1))
2245
2246 /*
2247 Top
2248
2249 The top-most available chunk (i.e., the one bordering the end of
2250 available memory) is treated specially. It is never included in
2251 any bin, is used only if no other chunk is available, and is
2252 released back to the system if it is very large (see
2253 M_TRIM_THRESHOLD). Because top initially
2254 points to its own bin with initial zero size, thus forcing
2255 extension on the first malloc request, we avoid having any special
2256 code in malloc to check whether it even exists yet. But we still
2257 need to do so when getting memory from system, so we make
2258 initial_top treat the bin as a legal but unusable chunk during the
2259 interval between initialization and the first call to
2260 sYSMALLOc. (This is somewhat delicate, since it relies on
2261 the 2 preceding words to be zero during this interval as well.)
2262 */
2263
2264 /* Conveniently, the unsorted bin can be used as dummy top on first call */
2265 #define initial_top(M) (unsorted_chunks(M))
2266
2267 /*
2268 Binmap
2269
2270 To help compensate for the large number of bins, a one-level index
2271 structure is used for bin-by-bin searching. `binmap' is a
2272 bitvector recording whether bins are definitely empty so they can
2273 be skipped over during during traversals. The bits are NOT always
2274 cleared as soon as bins are empty, but instead only
2275 when they are noticed to be empty during traversal in malloc.
2276 */
2277
2278 /* Conservatively use 32 bits per map word, even if on 64bit system */
2279 #define BINMAPSHIFT 5
2280 #define BITSPERMAP (1U << BINMAPSHIFT)
2281 #define BINMAPSIZE (NBINS / BITSPERMAP)
2282
2283 #define idx2block(i) ((i) >> BINMAPSHIFT)
2284 #define idx2bit(i) ((1U << ((i) & ((1U << BINMAPSHIFT)-1))))
2285
2286 #define mark_bin(m,i) ((m)->binmap[idx2block(i)] |= idx2bit(i))
2287 #define unmark_bin(m,i) ((m)->binmap[idx2block(i)] &= ~(idx2bit(i)))
2288 #define get_binmap(m,i) ((m)->binmap[idx2block(i)] & idx2bit(i))
2289
2290 /*
2291 Fastbins
2292
2293 An array of lists holding recently freed small chunks. Fastbins
2294 are not doubly linked. It is faster to single-link them, and
2295 since chunks are never removed from the middles of these lists,
2296 double linking is not necessary. Also, unlike regular bins, they
2297 are not even processed in FIFO order (they use faster LIFO) since
2298 ordering doesn't much matter in the transient contexts in which
2299 fastbins are normally used.
2300
2301 Chunks in fastbins keep their inuse bit set, so they cannot
2302 be consolidated with other free chunks. malloc_consolidate
2303 releases all chunks in fastbins and consolidates them with
2304 other free chunks.
2305 */
2306
2307 typedef struct malloc_chunk* mfastbinptr;
2308
2309 /* offset 2 to use otherwise unindexable first 2 bins */
2310 #define fastbin_index(sz) ((((unsigned int)(sz)) >> 3) - 2)
2311
2312 /* The maximum fastbin request size we support */
2313 #define MAX_FAST_SIZE 80
2314
2315 #define NFASTBINS (fastbin_index(request2size(MAX_FAST_SIZE))+1)
2316
2317 /*
2318 FASTBIN_CONSOLIDATION_THRESHOLD is the size of a chunk in free()
2319 that triggers automatic consolidation of possibly-surrounding
2320 fastbin chunks. This is a heuristic, so the exact value should not
2321 matter too much. It is defined at half the default trim threshold as a
2322 compromise heuristic to only attempt consolidation if it is likely
2323 to lead to trimming. However, it is not dynamically tunable, since
2324 consolidation reduces fragmentation surrounding loarge chunks even
2325 if trimming is not used.
2326 */
2327
2328 #define FASTBIN_CONSOLIDATION_THRESHOLD \
2329 ((unsigned long)(DEFAULT_TRIM_THRESHOLD) >> 1)
2330
2331 /*
2332 Since the lowest 2 bits in max_fast don't matter in size comparisons,
2333 they are used as flags.
2334 */
2335
2336 /*
2337 ANYCHUNKS_BIT held in max_fast indicates that there may be any
2338 freed chunks at all. It is set true when entering a chunk into any
2339 bin.
2340 */
2341
2342 #define ANYCHUNKS_BIT (1U)
2343
2344 #define have_anychunks(M) (((M)->max_fast & ANYCHUNKS_BIT))
2345 #define set_anychunks(M) ((M)->max_fast |= ANYCHUNKS_BIT)
2346 #define clear_anychunks(M) ((M)->max_fast &= ~ANYCHUNKS_BIT)
2347
2348 /*
2349 FASTCHUNKS_BIT held in max_fast indicates that there are probably
2350 some fastbin chunks. It is set true on entering a chunk into any
2351 fastbin, and cleared only in malloc_consolidate.
2352 */
2353
2354 #define FASTCHUNKS_BIT (2U)
2355
2356 #define have_fastchunks(M) (((M)->max_fast & FASTCHUNKS_BIT))
2357 #define set_fastchunks(M) ((M)->max_fast |= (FASTCHUNKS_BIT|ANYCHUNKS_BIT))
2358 #define clear_fastchunks(M) ((M)->max_fast &= ~(FASTCHUNKS_BIT))
2359
2360 /*
2361 Set value of max_fast.
2362 Use impossibly small value if 0.
2363 */
2364
2365 #define set_max_fast(M, s) \
2366 (M)->max_fast = (((s) == 0)? SMALLBIN_WIDTH: request2size(s)) | \
2367 ((M)->max_fast & (FASTCHUNKS_BIT|ANYCHUNKS_BIT))
2368
2369 #define get_max_fast(M) \
2370 ((M)->max_fast & ~(FASTCHUNKS_BIT | ANYCHUNKS_BIT))
2371
2372
2373 /*
2374 morecore_properties is a status word holding dynamically discovered
2375 or controlled properties of the morecore function
2376 */
2377
2378 #define MORECORE_CONTIGUOUS_BIT (1U)
2379
2380 #define contiguous(M) \
2381 (((M)->morecore_properties & MORECORE_CONTIGUOUS_BIT))
2382 #define noncontiguous(M) \
2383 (((M)->morecore_properties & MORECORE_CONTIGUOUS_BIT) == 0)
2384 #define set_contiguous(M) \
2385 ((M)->morecore_properties |= MORECORE_CONTIGUOUS_BIT)
2386 #define set_noncontiguous(M) \
2387 ((M)->morecore_properties &= ~MORECORE_CONTIGUOUS_BIT)
2388
2389
2390 /*
2391 ----------- Internal state representation and initialization -----------
2392 */
2393
2394 struct malloc_state {
2395
2396 /* The maximum chunk size to be eligible for fastbin */
2397 INTERNAL_SIZE_T max_fast; /* low 2 bits used as flags */
2398
2399 /* Fastbins */
2400 mfastbinptr fastbins[NFASTBINS];
2401
2402 /* Base of the topmost chunk -- not otherwise kept in a bin */
2403 mchunkptr top;
2404
2405 /* The remainder from the most recent split of a small request */
2406 mchunkptr last_remainder;
2407
2408 /* Normal bins packed as described above */
2409 mchunkptr bins[NBINS * 2];
2410
2411 /* Bitmap of bins. Trailing zero map handles cases of largest binned size */
2412 unsigned int binmap[BINMAPSIZE+1];
2413
2414 /* Tunable parameters */
2415 CHUNK_SIZE_T trim_threshold;
2416 INTERNAL_SIZE_T top_pad;
2417 INTERNAL_SIZE_T mmap_threshold;
2418
2419 /* Memory map support */
2420 int n_mmaps;
2421 int n_mmaps_max;
2422 int max_n_mmaps;
2423
2424 /* Cache malloc_getpagesize */
2425 unsigned int pagesize;
2426
2427 /* Track properties of MORECORE */
2428 unsigned int morecore_properties;
2429
2430 /* Statistics */
2431 INTERNAL_SIZE_T mmapped_mem;
2432 INTERNAL_SIZE_T sbrked_mem;
2433 INTERNAL_SIZE_T max_sbrked_mem;
2434 INTERNAL_SIZE_T max_mmapped_mem;
2435 INTERNAL_SIZE_T max_total_mem;
2436 };
2437
2438 typedef struct malloc_state *mstate;
2439
2440 /*
2441 There is exactly one instance of this struct in this malloc.
2442 If you are adapting this malloc in a way that does NOT use a static
2443 malloc_state, you MUST explicitly zero-fill it before using. This
2444 malloc relies on the property that malloc_state is initialized to
2445 all zeroes (as is true of C statics).
2446 */
2447
2448 static struct malloc_state av_; /* never directly referenced */
2449
2450 /*
2451 All uses of av_ are via get_malloc_state().
2452 At most one "call" to get_malloc_state is made per invocation of
2453 the public versions of malloc and free, but other routines
2454 that in turn invoke malloc and/or free may call more then once.
2455 Also, it is called in check* routines if DEBUG is set.
2456 */
2457
2458 #define get_malloc_state() (&(av_))
2459
2460 /*
2461 Initialize a malloc_state struct.
2462
2463 This is called only from within malloc_consolidate, which needs
2464 be called in the same contexts anyway. It is never called directly
2465 outside of malloc_consolidate because some optimizing compilers try
2466 to inline it at all call points, which turns out not to be an
2467 optimization at all. (Inlining it in malloc_consolidate is fine though.)
2468 */
2469
2470 #if __STD_C
2471 static void malloc_init_state(mstate av)
2472 #else
2473 static void malloc_init_state(av) mstate av;
2474 #endif
2475 {
2476 int i;
2477 mbinptr bin;
2478
2479 /* Establish circular links for normal bins */
2480 for (i = 1; i < NBINS; ++i) {
2481 bin = bin_at(av,i);
2482 bin->fd = bin->bk = bin;
2483 }
2484
2485 av->top_pad = DEFAULT_TOP_PAD;
2486 av->n_mmaps_max = DEFAULT_MMAP_MAX;
2487 av->mmap_threshold = DEFAULT_MMAP_THRESHOLD;
2488 av->trim_threshold = DEFAULT_TRIM_THRESHOLD;
2489
2490 #if MORECORE_CONTIGUOUS
2491 set_contiguous(av);
2492 #else
2493 set_noncontiguous(av);
2494 #endif
2495
2496
2497 set_max_fast(av, DEFAULT_MXFAST);
2498
2499 av->top = initial_top(av);
2500 av->pagesize = malloc_getpagesize;
2501 }
2502
2503 /*
2504 Other internal utilities operating on mstates
2505 */
2506
2507 #if __STD_C
2508 static Void_t* sYSMALLOc(INTERNAL_SIZE_T, mstate);
2509 static int sYSTRIm(size_t, mstate);
2510 static void malloc_consolidate(mstate);
2511 static Void_t** iALLOc(size_t, size_t*, int, Void_t**);
2512 #else
2513 static Void_t* sYSMALLOc();
2514 static int sYSTRIm();
2515 static void malloc_consolidate();
2516 static Void_t** iALLOc();
2517 #endif
2518
2519 /*
2520 Debugging support
2521
2522 These routines make a number of assertions about the states
2523 of data structures that should be true at all times. If any
2524 are not true, it's very likely that a user program has somehow
2525 trashed memory. (It's also possible that there is a coding error
2526 in malloc. In which case, please report it!)
2527 */
2528
2529 #if ! DEBUG
2530
2531 #define check_chunk(P)
2532 #define check_free_chunk(P)
2533 #define check_inuse_chunk(P)
2534 #define check_remalloced_chunk(P,N)
2535 #define check_malloced_chunk(P,N)
2536 #define check_malloc_state()
2537
2538 #else
2539 #define check_chunk(P) do_check_chunk(P)
2540 #define check_free_chunk(P) do_check_free_chunk(P)
2541 #define check_inuse_chunk(P) do_check_inuse_chunk(P)
2542 #define check_remalloced_chunk(P,N) do_check_remalloced_chunk(P,N)
2543 #define check_malloced_chunk(P,N) do_check_malloced_chunk(P,N)
2544 #define check_malloc_state() do_check_malloc_state()
2545
2546 /*
2547 Properties of all chunks
2548 */
2549
2550 #if __STD_C
2551 static void do_check_chunk(mchunkptr p)
2552 #else
2553 static void do_check_chunk(p) mchunkptr p;
2554 #endif
2555 {
2556 mstate av = get_malloc_state();
2557 CHUNK_SIZE_T sz = chunksize(p);
2558 /* min and max possible addresses assuming contiguous allocation */
2559 char* max_address = (char*)(av->top) + chunksize(av->top);
2560 char* min_address = max_address - av->sbrked_mem;
2561
2562 if (!chunk_is_mmapped(p)) {
2563
2564 /* Has legal address ... */
2565 if (p != av->top) {
2566 if (contiguous(av)) {
2567 assert(((char*)p) >= min_address);
2568 assert(((char*)p + sz) <= ((char*)(av->top)));
2569 }
2570 }
2571 else {
2572 /* top size is always at least MINSIZE */
2573 assert((CHUNK_SIZE_T)(sz) >= MINSIZE);
2574 /* top predecessor always marked inuse */
2575 assert(prev_inuse(p));
2576 }
2577
2578 }
2579 else {
2580 #if HAVE_MMAP
2581 /* address is outside main heap */
2582 if (contiguous(av) && av->top != initial_top(av)) {
2583 assert(((char*)p) < min_address || ((char*)p) > max_address);
2584 }
2585 /* chunk is page-aligned */
2586 assert(((p->prev_size + sz) & (av->pagesize-1)) == 0);
2587 /* mem is aligned */
2588 assert(aligned_OK(chunk2mem(p)));
2589 #else
2590 /* force an appropriate assert violation if debug set */
2591 assert(!chunk_is_mmapped(p));
2592 #endif
2593 }
2594 }
2595
2596 /*
2597 Properties of free chunks
2598 */
2599
2600 #if __STD_C
2601 static void do_check_free_chunk(mchunkptr p)
2602 #else
2603 static void do_check_free_chunk(p) mchunkptr p;
2604 #endif
2605 {
2606 mstate av = get_malloc_state();
2607
2608 INTERNAL_SIZE_T sz = p->size & ~PREV_INUSE;
2609 mchunkptr next = chunk_at_offset(p, sz);
2610
2611 do_check_chunk(p);
2612
2613 /* Chunk must claim to be free ... */
2614 assert(!inuse(p));
2615 assert (!chunk_is_mmapped(p));
2616
2617 /* Unless a special marker, must have OK fields */
2618 if ((CHUNK_SIZE_T)(sz) >= MINSIZE)
2619 {
2620 assert((sz & MALLOC_ALIGN_MASK) == 0);
2621 assert(aligned_OK(chunk2mem(p)));
2622 /* ... matching footer field */
2623 assert(next->prev_size == sz);
2624 /* ... and is fully consolidated */
2625 assert(prev_inuse(p));
2626 assert (next == av->top || inuse(next));
2627
2628 /* ... and has minimally sane links */
2629 assert(p->fd->bk == p);
2630 assert(p->bk->fd == p);
2631 }
2632 else /* markers are always of size SIZE_SZ */
2633 assert(sz == SIZE_SZ);
2634 }
2635
2636 /*
2637 Properties of inuse chunks
2638 */
2639
2640 #if __STD_C
2641 static void do_check_inuse_chunk(mchunkptr p)
2642 #else
2643 static void do_check_inuse_chunk(p) mchunkptr p;
2644 #endif
2645 {
2646 mstate av = get_malloc_state();
2647 mchunkptr next;
2648 do_check_chunk(p);
2649
2650 if (chunk_is_mmapped(p))
2651 return; /* mmapped chunks have no next/prev */
2652
2653 /* Check whether it claims to be in use ... */
2654 assert(inuse(p));
2655
2656 next = next_chunk(p);
2657
2658 /* ... and is surrounded by OK chunks.
2659 Since more things can be checked with free chunks than inuse ones,
2660 if an inuse chunk borders them and debug is on, it's worth doing them.
2661 */
2662 if (!prev_inuse(p)) {
2663 /* Note that we cannot even look at prev unless it is not inuse */
2664 mchunkptr prv = prev_chunk(p);
2665 assert(next_chunk(prv) == p);
2666 do_check_free_chunk(prv);
2667 }
2668
2669 if (next == av->top) {
2670 assert(prev_inuse(next));
2671 assert(chunksize(next) >= MINSIZE);
2672 }
2673 else if (!inuse(next))
2674 do_check_free_chunk(next);
2675 }
2676
2677 /*
2678 Properties of chunks recycled from fastbins
2679 */
2680
2681 #if __STD_C
2682 static void do_check_remalloced_chunk(mchunkptr p, INTERNAL_SIZE_T s)
2683 #else
2684 static void do_check_remalloced_chunk(p, s) mchunkptr p; INTERNAL_SIZE_T s;
2685 #endif
2686 {
2687 INTERNAL_SIZE_T sz = p->size & ~PREV_INUSE;
2688
2689 do_check_inuse_chunk(p);
2690
2691 /* Legal size ... */
2692 assert((sz & MALLOC_ALIGN_MASK) == 0);
2693 assert((CHUNK_SIZE_T)(sz) >= MINSIZE);
2694 /* ... and alignment */
2695 assert(aligned_OK(chunk2mem(p)));
2696 /* chunk is less than MINSIZE more than request */
2697 assert((long)(sz) - (long)(s) >= 0);
2698 assert((long)(sz) - (long)(s + MINSIZE) < 0);
2699 }
2700
2701 /*
2702 Properties of nonrecycled chunks at the point they are malloced
2703 */
2704
2705 #if __STD_C
2706 static void do_check_malloced_chunk(mchunkptr p, INTERNAL_SIZE_T s)
2707 #else
2708 static void do_check_malloced_chunk(p, s) mchunkptr p; INTERNAL_SIZE_T s;
2709 #endif
2710 {
2711 /* same as recycled case ... */
2712 do_check_remalloced_chunk(p, s);
2713
2714 /*
2715 ... plus, must obey implementation invariant that prev_inuse is
2716 always true of any allocated chunk; i.e., that each allocated
2717 chunk borders either a previously allocated and still in-use
2718 chunk, or the base of its memory arena. This is ensured
2719 by making all allocations from the the `lowest' part of any found
2720 chunk. This does not necessarily hold however for chunks
2721 recycled via fastbins.
2722 */
2723
2724 assert(prev_inuse(p));
2725 }
2726
2727
2728 /*
2729 Properties of malloc_state.
2730
2731 This may be useful for debugging malloc, as well as detecting user
2732 programmer errors that somehow write into malloc_state.
2733
2734 If you are extending or experimenting with this malloc, you can
2735 probably figure out how to hack this routine to print out or
2736 display chunk addresses, sizes, bins, and other instrumentation.
2737 */
2738
2739 static void do_check_malloc_state()
2740 {
2741 mstate av = get_malloc_state();
2742 int i;
2743 mchunkptr p;
2744 mchunkptr q;
2745 mbinptr b;
2746 unsigned int binbit;
2747 int empty;
2748 unsigned int idx;
2749 INTERNAL_SIZE_T size;
2750 CHUNK_SIZE_T total = 0;
2751 int max_fast_bin;
2752
2753 /* internal size_t must be no wider than pointer type */
2754 assert(sizeof(INTERNAL_SIZE_T) <= sizeof(char*));
2755
2756 /* alignment is a power of 2 */
2757 assert((MALLOC_ALIGNMENT & (MALLOC_ALIGNMENT-1)) == 0);
2758
2759 /* cannot run remaining checks until fully initialized */
2760 if (av->top == 0 || av->top == initial_top(av))
2761 return;
2762
2763 /* pagesize is a power of 2 */
2764 assert((av->pagesize & (av->pagesize-1)) == 0);
2765
2766 /* properties of fastbins */
2767
2768 /* max_fast is in allowed range */
2769 assert(get_max_fast(av) <= request2size(MAX_FAST_SIZE));
2770
2771 max_fast_bin = fastbin_index(av->max_fast);
2772
2773 for (i = 0; i < NFASTBINS; ++i) {
2774 p = av->fastbins[i];
2775
2776 /* all bins past max_fast are empty */
2777 if (i > max_fast_bin)
2778 assert(p == 0);
2779
2780 while (p != 0) {
2781 /* each chunk claims to be inuse */
2782 do_check_inuse_chunk(p);
2783 total += chunksize(p);
2784 /* chunk belongs in this bin */
2785 assert(fastbin_index(chunksize(p)) == i);
2786 p = p->fd;
2787 }
2788 }
2789
2790 if (total != 0)
2791 assert(have_fastchunks(av));
2792 else if (!have_fastchunks(av))
2793 assert(total == 0);
2794
2795 /* check normal bins */
2796 for (i = 1; i < NBINS; ++i) {
2797 b = bin_at(av,i);
2798
2799 /* binmap is accurate (except for bin 1 == unsorted_chunks) */
2800 if (i >= 2) {
2801 binbit = get_binmap(av,i);
2802 empty = last(b) == b;
2803 if (!binbit)
2804 assert(empty);
2805 else if (!empty)
2806 assert(binbit);
2807 }
2808
2809 for (p = last(b); p != b; p = p->bk) {
2810 /* each chunk claims to be free */
2811 do_check_free_chunk(p);
2812 size = chunksize(p);
2813 total += size;
2814 if (i >= 2) {
2815 /* chunk belongs in bin */
2816 idx = bin_index(size);
2817 assert(idx == i);
2818 /* lists are sorted */
2819 if ((CHUNK_SIZE_T) size >= (CHUNK_SIZE_T)(FIRST_SORTED_BIN_SIZE)) {
2820 assert(p->bk == b ||
2821 (CHUNK_SIZE_T)chunksize(p->bk) >=
2822 (CHUNK_SIZE_T)chunksize(p));
2823 }
2824 }
2825 /* chunk is followed by a legal chain of inuse chunks */
2826 for (q = next_chunk(p);
2827 (q != av->top && inuse(q) &&
2828 (CHUNK_SIZE_T)(chunksize(q)) >= MINSIZE);
2829 q = next_chunk(q))
2830 do_check_inuse_chunk(q);
2831 }
2832 }
2833
2834 /* top chunk is OK */
2835 check_chunk(av->top);
2836
2837 /* sanity checks for statistics */
2838
2839 assert(total <= (CHUNK_SIZE_T)(av->max_total_mem));
2840 assert(av->n_mmaps >= 0);
2841 assert(av->n_mmaps <= av->max_n_mmaps);
2842
2843 assert((CHUNK_SIZE_T)(av->sbrked_mem) <=
2844 (CHUNK_SIZE_T)(av->max_sbrked_mem));
2845
2846 assert((CHUNK_SIZE_T)(av->mmapped_mem) <=
2847 (CHUNK_SIZE_T)(av->max_mmapped_mem));
2848
2849 assert((CHUNK_SIZE_T)(av->max_total_mem) >=
2850 (CHUNK_SIZE_T)(av->mmapped_mem) + (CHUNK_SIZE_T)(av->sbrked_mem));
2851 }
2852 #endif
2853
2854
2855 /* ----------- Routines dealing with system allocation -------------- */
2856
2857 /*
2858 sysmalloc handles malloc cases requiring more memory from the system.
2859 On entry, it is assumed that av->top does not have enough
2860 space to service request for nb bytes, thus requiring that av->top
2861 be extended or replaced.
2862 */
2863
2864 #if __STD_C
2865 static Void_t* sYSMALLOc(INTERNAL_SIZE_T nb, mstate av)
2866 #else
2867 static Void_t* sYSMALLOc(nb, av) INTERNAL_SIZE_T nb; mstate av;
2868 #endif
2869 {
2870 mchunkptr old_top; /* incoming value of av->top */
2871 INTERNAL_SIZE_T old_size; /* its size */
2872 char* old_end; /* its end address */
2873
2874 long size; /* arg to first MORECORE or mmap call */
2875 char* brk; /* return value from MORECORE */
2876
2877 long correction; /* arg to 2nd MORECORE call */
2878 char* snd_brk; /* 2nd return val */
2879
2880 INTERNAL_SIZE_T front_misalign; /* unusable bytes at front of new space */
2881 INTERNAL_SIZE_T end_misalign; /* partial page left at end of new space */
2882 char* aligned_brk; /* aligned offset into brk */
2883
2884 mchunkptr p; /* the allocated/returned chunk */
2885 mchunkptr remainder; /* remainder from allocation */
2886 CHUNK_SIZE_T remainder_size; /* its size */
2887
2888 CHUNK_SIZE_T sum; /* for updating stats */
2889
2890 size_t pagemask = av->pagesize - 1;
2891
2892 /*
2893 If there is space available in fastbins, consolidate and retry
2894 malloc from scratch rather than getting memory from system. This
2895 can occur only if nb is in smallbin range so we didn't consolidate
2896 upon entry to malloc. It is much easier to handle this case here
2897 than in malloc proper.
2898 */
2899
2900 if (have_fastchunks(av)) {
2901 assert(in_smallbin_range(nb));
2902 malloc_consolidate(av);
2903 return mALLOc(nb - MALLOC_ALIGN_MASK);
2904 }
2905
2906
2907 #if HAVE_MMAP
2908
2909 /*
2910 If have mmap, and the request size meets the mmap threshold, and
2911 the system supports mmap, and there are few enough currently
2912 allocated mmapped regions, try to directly map this request
2913 rather than expanding top.
2914 */
2915
2916 if ((CHUNK_SIZE_T)(nb) >= (CHUNK_SIZE_T)(av->mmap_threshold) &&
2917 (av->n_mmaps < av->n_mmaps_max)) {
2918
2919 char* mm; /* return value from mmap call*/
2920
2921 /*
2922 Round up size to nearest page. For mmapped chunks, the overhead
2923 is one SIZE_SZ unit larger than for normal chunks, because there
2924 is no following chunk whose prev_size field could be used.
2925 */
2926 size = (nb + SIZE_SZ + MALLOC_ALIGN_MASK + pagemask) & ~pagemask;
2927
2928 /* Don't try if size wraps around 0 */
2929 if ((CHUNK_SIZE_T)(size) > (CHUNK_SIZE_T)(nb)) {
2930
2931 mm = (char*)(MMAP(0, size, PROT_READ|PROT_WRITE, MAP_PRIVATE));
2932
2933 if (mm != (char*)(MORECORE_FAILURE)) {
2934
2935 /*
2936 The offset to the start of the mmapped region is stored
2937 in the prev_size field of the chunk. This allows us to adjust
2938 returned start address to meet alignment requirements here
2939 and in memalign(), and still be able to compute proper
2940 address argument for later munmap in free() and realloc().
2941 */
2942
2943 front_misalign = (INTERNAL_SIZE_T)chunk2mem(mm) & MALLOC_ALIGN_MASK;
2944 if (front_misalign > 0) {
2945 correction = MALLOC_ALIGNMENT - front_misalign;
2946 p = (mchunkptr)(mm + correction);
2947 p->prev_size = correction;
2948 set_head(p, (size - correction) |IS_MMAPPED);
2949 }
2950 else {
2951 p = (mchunkptr)mm;
2952 p->prev_size = 0;
2953 set_head(p, size|IS_MMAPPED);
2954 }
2955
2956 /* update statistics */
2957
2958 if (++av->n_mmaps > av->max_n_mmaps)
2959 av->max_n_mmaps = av->n_mmaps;
2960
2961 sum = av->mmapped_mem += size;
2962 if (sum > (CHUNK_SIZE_T)(av->max_mmapped_mem))
2963 av->max_mmapped_mem = sum;
2964 sum += av->sbrked_mem;
2965 if (sum > (CHUNK_SIZE_T)(av->max_total_mem))
2966 av->max_total_mem = sum;
2967
2968 check_chunk(p);
2969
2970 return chunk2mem(p);
2971 }
2972 }
2973 }
2974 #endif
2975
2976 /* Record incoming configuration of top */
2977
2978 old_top = av->top;
2979 old_size = chunksize(old_top);
2980 old_end = (char*)(chunk_at_offset(old_top, old_size));
2981
2982 brk = snd_brk = (char*)(MORECORE_FAILURE);
2983
2984 /*
2985 If not the first time through, we require old_size to be
2986 at least MINSIZE and to have prev_inuse set.
2987 */
2988
2989 assert((old_top == initial_top(av) && old_size == 0) ||
2990 ((CHUNK_SIZE_T) (old_size) >= MINSIZE &&
2991 prev_inuse(old_top)));
2992
2993 /* Precondition: not enough current space to satisfy nb request */
2994 assert((CHUNK_SIZE_T)(old_size) < (CHUNK_SIZE_T)(nb + MINSIZE));
2995
2996 /* Precondition: all fastbins are consolidated */
2997 assert(!have_fastchunks(av));
2998
2999
3000 /* Request enough space for nb + pad + overhead */
3001
3002 size = nb + av->top_pad + MINSIZE;
3003
3004 /*
3005 If contiguous, we can subtract out existing space that we hope to
3006 combine with new space. We add it back later only if
3007 we don't actually get contiguous space.
3008 */
3009
3010 if (contiguous(av))
3011 size -= old_size;
3012
3013 /*
3014 Round to a multiple of page size.
3015 If MORECORE is not contiguous, this ensures that we only call it
3016 with whole-page arguments. And if MORECORE is contiguous and
3017 this is not first time through, this preserves page-alignment of
3018 previous calls. Otherwise, we correct to page-align below.
3019 */
3020
3021 size = (size + pagemask) & ~pagemask;
3022
3023 /*
3024 Don't try to call MORECORE if argument is so big as to appear
3025 negative. Note that since mmap takes size_t arg, it may succeed
3026 below even if we cannot call MORECORE.
3027 */
3028
3029 if (size > 0)
3030 brk = (char*)(MORECORE(size));
3031
3032 /*
3033 If have mmap, try using it as a backup when MORECORE fails or
3034 cannot be used. This is worth doing on systems that have "holes" in
3035 address space, so sbrk cannot extend to give contiguous space, but
3036 space is available elsewhere. Note that we ignore mmap max count
3037 and threshold limits, since the space will not be used as a
3038 segregated mmap region.
3039 */
3040
3041 #if HAVE_MMAP
3042 if (brk == (char*)(MORECORE_FAILURE)) {
3043
3044 /* Cannot merge with old top, so add its size back in */
3045 if (contiguous(av))
3046 size = (size + old_size + pagemask) & ~pagemask;
3047
3048 /* If we are relying on mmap as backup, then use larger units */
3049 if ((CHUNK_SIZE_T)(size) < (CHUNK_SIZE_T)(MMAP_AS_MORECORE_SIZE))
3050 size = MMAP_AS_MORECORE_SIZE;
3051
3052 /* Don't try if size wraps around 0 */
3053 if ((CHUNK_SIZE_T)(size) > (CHUNK_SIZE_T)(nb)) {
3054
3055 brk = (char*)(MMAP(0, size, PROT_READ|PROT_WRITE, MAP_PRIVATE));
3056
3057 if (brk != (char*)(MORECORE_FAILURE)) {
3058
3059 /* We do not need, and cannot use, another sbrk call to find end */
3060 snd_brk = brk + size;
3061
3062 /*
3063 Record that we no longer have a contiguous sbrk region.
3064 After the first time mmap is used as backup, we do not
3065 ever rely on contiguous space since this could incorrectly
3066 bridge regions.
3067 */
3068 set_noncontiguous(av);
3069 }
3070 }
3071 }
3072 #endif
3073
3074 if (brk != (char*)(MORECORE_FAILURE)) {
3075 av->sbrked_mem += size;
3076
3077 /*
3078 If MORECORE extends previous space, we can likewise extend top size.
3079 */
3080
3081 if (brk == old_end && snd_brk == (char*)(MORECORE_FAILURE)) {
3082 set_head(old_top, (size + old_size) | PREV_INUSE);
3083 }
3084
3085 /*
3086 Otherwise, make adjustments:
3087
3088 * If the first time through or noncontiguous, we need to call sbrk
3089 just to find out where the end of memory lies.
3090
3091 * We need to ensure that all returned chunks from malloc will meet
3092 MALLOC_ALIGNMENT
3093
3094 * If there was an intervening foreign sbrk, we need to adjust sbrk
3095 request size to account for fact that we will not be able to
3096 combine new space with existing space in old_top.
3097
3098 * Almost all systems internally allocate whole pages at a time, in
3099 which case we might as well use the whole last page of request.
3100 So we allocate enough more memory to hit a page boundary now,
3101 which in turn causes future contiguous calls to page-align.
3102 */
3103
3104 else {
3105 front_misalign = 0;
3106 end_misalign = 0;
3107 correction = 0;
3108 aligned_brk = brk;
3109
3110 /*
3111 If MORECORE returns an address lower than we have seen before,
3112 we know it isn't really contiguous. This and some subsequent
3113 checks help cope with non-conforming MORECORE functions and
3114 the presence of "foreign" calls to MORECORE from outside of
3115 malloc or by other threads. We cannot guarantee to detect
3116 these in all cases, but cope with the ones we do detect.
3117 */
3118 if (contiguous(av) && old_size != 0 && brk < old_end) {
3119 set_noncontiguous(av);
3120 }
3121
3122 /* handle contiguous cases */
3123 if (contiguous(av)) {
3124
3125 /*
3126 We can tolerate forward non-contiguities here (usually due
3127 to foreign calls) but treat them as part of our space for
3128 stats reporting.
3129 */
3130 if (old_size != 0)
3131 av->sbrked_mem += brk - old_end;
3132
3133 /* Guarantee alignment of first new chunk made from this space */
3134
3135 front_misalign = (INTERNAL_SIZE_T)chunk2mem(brk) & MALLOC_ALIGN_MASK;
3136 if (front_misalign > 0) {
3137
3138 /*
3139 Skip over some bytes to arrive at an aligned position.
3140 We don't need to specially mark these wasted front bytes.
3141 They will never be accessed anyway because
3142 prev_inuse of av->top (and any chunk created from its start)
3143 is always true after initialization.
3144 */
3145
3146 correction = MALLOC_ALIGNMENT - front_misalign;
3147 aligned_brk += correction;
3148 }
3149
3150 /*
3151 If this isn't adjacent to existing space, then we will not
3152 be able to merge with old_top space, so must add to 2nd request.
3153 */
3154
3155 correction += old_size;
3156
3157 /* Extend the end address to hit a page boundary */
3158 end_misalign = (INTERNAL_SIZE_T)(brk + size + correction);
3159 correction += ((end_misalign + pagemask) & ~pagemask) - end_misalign;
3160
3161 assert(correction >= 0);
3162 snd_brk = (char*)(MORECORE(correction));
3163
3164 if (snd_brk == (char*)(MORECORE_FAILURE)) {
3165 /*
3166 If can't allocate correction, try to at least find out current
3167 brk. It might be enough to proceed without failing.
3168 */
3169 correction = 0;
3170 snd_brk = (char*)(MORECORE(0));
3171 }
3172 else if (snd_brk < brk) {
3173 /*
3174 If the second call gives noncontiguous space even though
3175 it says it won't, the only course of action is to ignore
3176 results of second call, and conservatively estimate where
3177 the first call left us. Also set noncontiguous, so this
3178 won't happen again, leaving at most one hole.
3179
3180 Note that this check is intrinsically incomplete. Because
3181 MORECORE is allowed to give more space than we ask for,
3182 there is no reliable way to detect a noncontiguity
3183 producing a forward gap for the second call.
3184 */
3185 snd_brk = brk + size;
3186 correction = 0;
3187 set_noncontiguous(av);
3188 }
3189
3190 }
3191
3192 /* handle non-contiguous cases */
3193 else {
3194 /* MORECORE/mmap must correctly align */
3195 assert(aligned_OK(chunk2mem(brk)));
3196
3197 /* Find out current end of memory */
3198 if (snd_brk == (char*)(MORECORE_FAILURE)) {
3199 snd_brk = (char*)(MORECORE(0));
3200 av->sbrked_mem += snd_brk - brk - size;
3201 }
3202 }
3203
3204 /* Adjust top based on results of second sbrk */
3205 if (snd_brk != (char*)(MORECORE_FAILURE)) {
3206 av->top = (mchunkptr)aligned_brk;
3207 set_head(av->top, (snd_brk - aligned_brk + correction) | PREV_INUSE);
3208 av->sbrked_mem += correction;
3209
3210 /*
3211 If not the first time through, we either have a
3212 gap due to foreign sbrk or a non-contiguous region. Insert a
3213 double fencepost at old_top to prevent consolidation with space
3214 we don't own. These fenceposts are artificial chunks that are
3215 marked as inuse and are in any case too small to use. We need
3216 two to make sizes and alignments work out.
3217 */
3218
3219 if (old_size != 0) {
3220 /*
3221 Shrink old_top to insert fenceposts, keeping size a
3222 multiple of MALLOC_ALIGNMENT. We know there is at least
3223 enough space in old_top to do this.
3224 */
3225 old_size = (old_size - 3*SIZE_SZ) & ~MALLOC_ALIGN_MASK;
3226 set_head(old_top, old_size | PREV_INUSE);
3227
3228 /*
3229 Note that the following assignments completely overwrite
3230 old_top when old_size was previously MINSIZE. This is
3231 intentional. We need the fencepost, even if old_top otherwise gets
3232 lost.
3233 */
3234 chunk_at_offset(old_top, old_size )->size =
3235 SIZE_SZ|PREV_INUSE;
3236
3237 chunk_at_offset(old_top, old_size + SIZE_SZ)->size =
3238 SIZE_SZ|PREV_INUSE;
3239
3240 /*
3241 If possible, release the rest, suppressing trimming.
3242 */
3243 if (old_size >= MINSIZE) {
3244 INTERNAL_SIZE_T tt = av->trim_threshold;
3245 av->trim_threshold = (INTERNAL_SIZE_T)(-1);
3246 fREe(chunk2mem(old_top));
3247 av->trim_threshold = tt;
3248 }
3249 }
3250 }
3251 }
3252
3253 /* Update statistics */
3254 sum = av->sbrked_mem;
3255 if (sum > (CHUNK_SIZE_T)(av->max_sbrked_mem))
3256 av->max_sbrked_mem = sum;
3257
3258 sum += av->mmapped_mem;
3259 if (sum > (CHUNK_SIZE_T)(av->max_total_mem))
3260 av->max_total_mem = sum;
3261
3262 check_malloc_state();
3263
3264 /* finally, do the allocation */
3265
3266 p = av->top;
3267 size = chunksize(p);
3268
3269 /* check that one of the above allocation paths succeeded */
3270 if ((CHUNK_SIZE_T)(size) >= (CHUNK_SIZE_T)(nb + MINSIZE)) {
3271 remainder_size = size - nb;
3272 remainder = chunk_at_offset(p, nb);
3273 av->top = remainder;
3274 set_head(p, nb | PREV_INUSE);
3275 set_head(remainder, remainder_size | PREV_INUSE);
3276 check_malloced_chunk(p, nb);
3277 return chunk2mem(p);
3278 }
3279
3280 }
3281
3282 /* catch all failure paths */
3283 MALLOC_FAILURE_ACTION;
3284 return 0;
3285 }
3286
3287
3288
3289
3290 /*
3291 sYSTRIm is an inverse of sorts to sYSMALLOc. It gives memory back
3292 to the system (via negative arguments to sbrk) if there is unused
3293 memory at the `high' end of the malloc pool. It is called
3294 automatically by free() when top space exceeds the trim
3295 threshold. It is also called by the public malloc_trim routine. It
3296 returns 1 if it actually released any memory, else 0.
3297 */
3298
3299 #if __STD_C
3300 static int sYSTRIm(size_t pad, mstate av)
3301 #else
3302 static int sYSTRIm(pad, av) size_t pad; mstate av;
3303 #endif
3304 {
3305 long top_size; /* Amount of top-most memory */
3306 long extra; /* Amount to release */
3307 long released; /* Amount actually released */
3308 char* current_brk; /* address returned by pre-check sbrk call */
3309 char* new_brk; /* address returned by post-check sbrk call */
3310 size_t pagesz;
3311
3312 pagesz = av->pagesize;
3313 top_size = chunksize(av->top);
3314
3315 /* Release in pagesize units, keeping at least one page */
3316 extra = ((top_size - pad - MINSIZE + (pagesz-1)) / pagesz - 1) * pagesz;
3317
3318 if (extra > 0) {
3319
3320 /*
3321 Only proceed if end of memory is where we last set it.
3322 This avoids problems if there were foreign sbrk calls.
3323 */
3324 current_brk = (char*)(MORECORE(0));
3325 if (current_brk == (char*)(av->top) + top_size) {
3326
3327 /*
3328 Attempt to release memory. We ignore MORECORE return value,
3329 and instead call again to find out where new end of memory is.
3330 This avoids problems if first call releases less than we asked,
3331 of if failure somehow altered brk value. (We could still
3332 encounter problems if it altered brk in some very bad way,
3333 but the only thing we can do is adjust anyway, which will cause
3334 some downstream failure.)
3335 */
3336
3337 MORECORE(-extra);
3338 new_brk = (char*)(MORECORE(0));
3339
3340 if (new_brk != (char*)MORECORE_FAILURE) {
3341 released = (long)(current_brk - new_brk);
3342
3343 if (released != 0) {
3344 /* Success. Adjust top. */
3345 av->sbrked_mem -= released;
3346 set_head(av->top, (top_size - released) | PREV_INUSE);
3347 check_malloc_state();
3348 return 1;
3349 }
3350 }
3351 }
3352 }
3353 return 0;
3354 }
3355
3356 /* ================ MPR ================== */
3357
3358 static int __mpr_malloc_initialized = 0;
3359 Void_t* (*__mpr_malloc_hook)(size_t) = 0;
3360 void (*__mpr_free_hook)(Void_t*) = 0;
3361 Void_t* (*__mpr_realloc_hook)(Void_t*, size_t) = 0;
3362
3363 static void
3364 __mpr_malloc_initialize()
3365 {
3366 extern int mprhook();
3367 mprhook();
3368 __mpr_malloc_initialized = 1;
3369 }
3370
3371 /*
3372 ------------------------------ malloc ------------------------------
3373 */
3374
3375
3376 #if __STD_C
3377 Void_t* mALLOc(size_t bytes)
3378 #else
3379 Void_t* mALLOc(bytes) size_t bytes;
3380 #endif
3381 {
3382 mstate av = get_malloc_state();
3383
3384 INTERNAL_SIZE_T nb; /* normalized request size */
3385 unsigned int idx; /* associated bin index */
3386 mbinptr bin; /* associated bin */
3387 mfastbinptr* fb; /* associated fastbin */
3388
3389 mchunkptr victim; /* inspected/selected chunk */
3390 INTERNAL_SIZE_T size; /* its size */
3391 int victim_index; /* its bin index */
3392
3393 mchunkptr remainder; /* remainder from a split */
3394 CHUNK_SIZE_T remainder_size; /* its size */
3395
3396 unsigned int block; /* bit map traverser */
3397 unsigned int bit; /* bit map traverser */
3398 unsigned int map; /* current word of binmap */
3399
3400 mchunkptr fwd; /* misc temp for linking */
3401 mchunkptr bck; /* misc temp for linking */
3402
3403 if (!__mpr_malloc_initialized)
3404 __mpr_malloc_initialize();
3405 if (__mpr_malloc_hook != NULL)
3406 return (*__mpr_malloc_hook)(bytes);
3407
3408 /*
3409 Convert request size to internal form by adding SIZE_SZ bytes
3410 overhead plus possibly more to obtain necessary alignment and/or
3411 to obtain a size of at least MINSIZE, the smallest allocatable
3412 size. Also, checked_request2size traps (returning 0) request sizes
3413 that are so large that they wrap around zero when padded and
3414 aligned.
3415 */
3416
3417 checked_request2size(bytes, nb);
3418
3419 /*
3420 Bypass search if no frees yet
3421 */
3422 if (!have_anychunks(av)) {
3423 if (av->max_fast == 0) /* initialization check */
3424 malloc_consolidate(av);
3425 goto use_top;
3426 }
3427
3428 /*
3429 If the size qualifies as a fastbin, first check corresponding bin.
3430 */
3431
3432 if ((CHUNK_SIZE_T)(nb) <= (CHUNK_SIZE_T)(av->max_fast)) {
3433 fb = &(av->fastbins[(fastbin_index(nb))]);
3434 if ( (victim = *fb) != 0) {
3435 *fb = victim->fd;
3436 check_remalloced_chunk(victim, nb);
3437 return chunk2mem(victim);
3438 }
3439 }
3440
3441 /*
3442 If a small request, check regular bin. Since these "smallbins"
3443 hold one size each, no searching within bins is necessary.
3444 (For a large request, we need to wait until unsorted chunks are
3445 processed to find best fit. But for small ones, fits are exact
3446 anyway, so we can check now, which is faster.)
3447 */
3448
3449 if (in_smallbin_range(nb)) {
3450 idx = smallbin_index(nb);
3451 bin = bin_at(av,idx);
3452
3453 if ( (victim = last(bin)) != bin) {
3454 bck = victim->bk;
3455 set_inuse_bit_at_offset(victim, nb);
3456 bin->bk = bck;
3457 bck->fd = bin;
3458
3459 check_malloced_chunk(victim, nb);
3460 return chunk2mem(victim);
3461 }
3462 }
3463
3464 /*
3465 If this is a large request, consolidate fastbins before continuing.
3466 While it might look excessive to kill all fastbins before
3467 even seeing if there is space available, this avoids
3468 fragmentation problems normally associated with fastbins.
3469 Also, in practice, programs tend to have runs of either small or
3470 large requests, but less often mixtures, so consolidation is not
3471 invoked all that often in most programs. And the programs that
3472 it is called frequently in otherwise tend to fragment.
3473 */
3474
3475 else {
3476 idx = largebin_index(nb);
3477 if (have_fastchunks(av))
3478 malloc_consolidate(av);
3479 }
3480
3481 /*
3482 Process recently freed or remaindered chunks, taking one only if
3483 it is exact fit, or, if this a small request, the chunk is remainder from
3484 the most recent non-exact fit. Place other traversed chunks in
3485 bins. Note that this step is the only place in any routine where
3486 chunks are placed in bins.
3487 */
3488
3489 while ( (victim = unsorted_chunks(av)->bk) != unsorted_chunks(av)) {
3490 bck = victim->bk;
3491 size = chunksize(victim);
3492
3493 /*
3494 If a small request, try to use last remainder if it is the
3495 only chunk in unsorted bin. This helps promote locality for
3496 runs of consecutive small requests. This is the only
3497 exception to best-fit, and applies only when there is
3498 no exact fit for a small chunk.
3499 */
3500
3501 if (in_smallbin_range(nb) &&
3502 bck == unsorted_chunks(av) &&
3503 victim == av->last_remainder &&
3504 (CHUNK_SIZE_T)(size) > (CHUNK_SIZE_T)(nb + MINSIZE)) {
3505
3506 /* split and reattach remainder */
3507 remainder_size = size - nb;
3508 remainder = chunk_at_offset(victim, nb);
3509 unsorted_chunks(av)->bk = unsorted_chunks(av)->fd = remainder;
3510 av->last_remainder = remainder;
3511 remainder->bk = remainder->fd = unsorted_chunks(av);
3512
3513 set_head(victim, nb | PREV_INUSE);
3514 set_head(remainder, remainder_size | PREV_INUSE);
3515 set_foot(remainder, remainder_size);
3516
3517 check_malloced_chunk(victim, nb);
3518 return chunk2mem(victim);
3519 }
3520
3521 /* remove from unsorted list */
3522 unsorted_chunks(av)->bk = bck;
3523 bck->fd = unsorted_chunks(av);
3524
3525 /* Take now instead of binning if exact fit */
3526
3527 if (size == nb) {
3528 set_inuse_bit_at_offset(victim, size);
3529 check_malloced_chunk(victim, nb);
3530 return chunk2mem(victim);
3531 }
3532
3533 /* place chunk in bin */
3534
3535 if (in_smallbin_range(size)) {
3536 victim_index = smallbin_index(size);
3537 bck = bin_at(av, victim_index);
3538 fwd = bck->fd;
3539 }
3540 else {
3541 victim_index = largebin_index(size);
3542 bck = bin_at(av, victim_index);
3543 fwd = bck->fd;
3544
3545 if (fwd != bck) {
3546 /* if smaller than smallest, place first */
3547 if ((CHUNK_SIZE_T)(size) < (CHUNK_SIZE_T)(bck->bk->size)) {
3548 fwd = bck;
3549 bck = bck->bk;
3550 }
3551 else if ((CHUNK_SIZE_T)(size) >=
3552 (CHUNK_SIZE_T)(FIRST_SORTED_BIN_SIZE)) {
3553
3554 /* maintain large bins in sorted order */
3555 size |= PREV_INUSE; /* Or with inuse bit to speed comparisons */
3556 while ((CHUNK_SIZE_T)(size) < (CHUNK_SIZE_T)(fwd->size))
3557 fwd = fwd->fd;
3558 bck = fwd->bk;
3559 }
3560 }
3561 }
3562
3563 mark_bin(av, victim_index);
3564 victim->bk = bck;
3565 victim->fd = fwd;
3566 fwd->bk = victim;
3567 bck->fd = victim;
3568 }
3569
3570 /*
3571 If a large request, scan through the chunks of current bin to
3572 find one that fits. (This will be the smallest that fits unless
3573 FIRST_SORTED_BIN_SIZE has been changed from default.) This is
3574 the only step where an unbounded number of chunks might be
3575 scanned without doing anything useful with them. However the
3576 lists tend to be short.
3577 */
3578
3579 if (!in_smallbin_range(nb)) {
3580 bin = bin_at(av, idx);
3581
3582 for (victim = last(bin); victim != bin; victim = victim->bk) {
3583 size = chunksize(victim);
3584
3585 if ((CHUNK_SIZE_T)(size) >= (CHUNK_SIZE_T)(nb)) {
3586 remainder_size = size - nb;
3587 unlink(victim, bck, fwd);
3588
3589 /* Exhaust */
3590 if (remainder_size < MINSIZE) {
3591 set_inuse_bit_at_offset(victim, size);
3592 check_malloced_chunk(victim, nb);
3593 return chunk2mem(victim);
3594 }
3595 /* Split */
3596 else {
3597 remainder = chunk_at_offset(victim, nb);
3598 unsorted_chunks(av)->bk = unsorted_chunks(av)->fd = remainder;
3599 remainder->bk = remainder->fd = unsorted_chunks(av);
3600 set_head(victim, nb | PREV_INUSE);
3601 set_head(remainder, remainder_size | PREV_INUSE);
3602 set_foot(remainder, remainder_size);
3603 check_malloced_chunk(victim, nb);
3604 return chunk2mem(victim);
3605 }
3606 }
3607 }
3608 }
3609
3610 /*
3611 Search for a chunk by scanning bins, starting with next largest
3612 bin. This search is strictly by best-fit; i.e., the smallest
3613 (with ties going to approximately the least recently used) chunk
3614 that fits is selected.
3615
3616 The bitmap avoids needing to check that most blocks are nonempty.
3617 */
3618
3619 ++idx;
3620 bin = bin_at(av,idx);
3621 block = idx2block(idx);
3622 map = av->binmap[block];
3623 bit = idx2bit(idx);
3624
3625 for (;;) {
3626
3627 /* Skip rest of block if there are no more set bits in this block. */
3628 if (bit > map || bit == 0) {
3629 do {
3630 if (++block >= BINMAPSIZE) /* out of bins */
3631 goto use_top;
3632 } while ( (map = av->binmap[block]) == 0);
3633
3634 bin = bin_at(av, (block << BINMAPSHIFT));
3635 bit = 1;
3636 }
3637
3638 /* Advance to bin with set bit. There must be one. */
3639 while ((bit & map) == 0) {
3640 bin = next_bin(bin);
3641 bit <<= 1;
3642 assert(bit != 0);
3643 }
3644
3645 /* Inspect the bin. It is likely to be non-empty */
3646 victim = last(bin);
3647
3648 /* If a false alarm (empty bin), clear the bit. */
3649 if (victim == bin) {
3650 av->binmap[block] = map &= ~bit; /* Write through */
3651 bin = next_bin(bin);
3652 bit <<= 1;
3653 }
3654
3655 else {
3656 size = chunksize(victim);
3657
3658 /* We know the first chunk in this bin is big enough to use. */
3659 assert((CHUNK_SIZE_T)(size) >= (CHUNK_SIZE_T)(nb));
3660
3661 remainder_size = size - nb;
3662
3663 /* unlink */
3664 bck = victim->bk;
3665 bin->bk = bck;
3666 bck->fd = bin;
3667
3668 /* Exhaust */
3669 if (remainder_size < MINSIZE) {
3670 set_inuse_bit_at_offset(victim, size);
3671 check_malloced_chunk(victim, nb);
3672 return chunk2mem(victim);
3673 }
3674
3675 /* Split */
3676 else {
3677 remainder = chunk_at_offset(victim, nb);
3678
3679 unsorted_chunks(av)->bk = unsorted_chunks(av)->fd = remainder;
3680 remainder->bk = remainder->fd = unsorted_chunks(av);
3681 /* advertise as last remainder */
3682 if (in_smallbin_range(nb))
3683 av->last_remainder = remainder;
3684
3685 set_head(victim, nb | PREV_INUSE);
3686 set_head(remainder, remainder_size | PREV_INUSE);
3687 set_foot(remainder, remainder_size);
3688 check_malloced_chunk(victim, nb);
3689 return chunk2mem(victim);
3690 }
3691 }
3692 }
3693
3694 use_top:
3695 /*
3696 If large enough, split off the chunk bordering the end of memory
3697 (held in av->top). Note that this is in accord with the best-fit
3698 search rule. In effect, av->top is treated as larger (and thus
3699 less well fitting) than any other available chunk since it can
3700 be extended to be as large as necessary (up to system
3701 limitations).
3702
3703 We require that av->top always exists (i.e., has size >=
3704 MINSIZE) after initialization, so if it would otherwise be
3705 exhuasted by current request, it is replenished. (The main
3706 reason for ensuring it exists is that we may need MINSIZE space
3707 to put in fenceposts in sysmalloc.)
3708 */
3709
3710 victim = av->top;
3711 size = chunksize(victim);
3712
3713 if ((CHUNK_SIZE_T)(size) >= (CHUNK_SIZE_T)(nb + MINSIZE)) {
3714 remainder_size = size - nb;
3715 remainder = chunk_at_offset(victim, nb);
3716 av->top = remainder;
3717 set_head(victim, nb | PREV_INUSE);
3718 set_head(remainder, remainder_size | PREV_INUSE);
3719
3720 check_malloced_chunk(victim, nb);
3721 return chunk2mem(victim);
3722 }
3723
3724 /*
3725 If no space in top, relay to handle system-dependent cases
3726 */
3727 return sYSMALLOc(nb, av);
3728 }
3729
3730 /*
3731 ------------------------------ free ------------------------------
3732 */
3733
3734 #if __STD_C
3735 void fREe(Void_t* mem)
3736 #else
3737 void fREe(mem) Void_t* mem;
3738 #endif
3739 {
3740 mstate av = get_malloc_state();
3741
3742 mchunkptr p; /* chunk corresponding to mem */
3743 INTERNAL_SIZE_T size; /* its size */
3744 mfastbinptr* fb; /* associated fastbin */
3745 mchunkptr nextchunk; /* next contiguous chunk */
3746 INTERNAL_SIZE_T nextsize; /* its size */
3747 int nextinuse; /* true if nextchunk is used */
3748 INTERNAL_SIZE_T prevsize; /* size of previous contiguous chunk */
3749 mchunkptr bck; /* misc temp for linking */
3750 mchunkptr fwd; /* misc temp for linking */
3751
3752 if (!__mpr_malloc_initialized)
3753 __mpr_malloc_initialize();
3754 if (__mpr_free_hook != NULL) {
3755 (*__mpr_free_hook)(mem);
3756 return;
3757 }
3758
3759 /* free(0) has no effect */
3760 if (mem != 0) {
3761 p = mem2chunk(mem);
3762 size = chunksize(p);
3763
3764 check_inuse_chunk(p);
3765
3766 /*
3767 If eligible, place chunk on a fastbin so it can be found
3768 and used quickly in malloc.
3769 */
3770
3771 if ((CHUNK_SIZE_T)(size) <= (CHUNK_SIZE_T)(av->max_fast)
3772
3773 #if TRIM_FASTBINS
3774 /*
3775 If TRIM_FASTBINS set, don't place chunks
3776 bordering top into fastbins
3777 */
3778 && (chunk_at_offset(p, size) != av->top)
3779 #endif
3780 ) {
3781
3782 set_fastchunks(av);
3783 fb = &(av->fastbins[fastbin_index(size)]);
3784 p->fd = *fb;
3785 *fb = p;
3786 }
3787
3788 /*
3789 Consolidate other non-mmapped chunks as they arrive.
3790 */
3791
3792 else if (!chunk_is_mmapped(p)) {
3793 set_anychunks(av);
3794
3795 nextchunk = chunk_at_offset(p, size);
3796 nextsize = chunksize(nextchunk);
3797
3798 /* consolidate backward */
3799 if (!prev_inuse(p)) {
3800 prevsize = p->prev_size;
3801 size += prevsize;
3802 p = chunk_at_offset(p, -((long) prevsize));
3803 unlink(p, bck, fwd);
3804 }
3805
3806 if (nextchunk != av->top) {
3807 /* get and clear inuse bit */
3808 nextinuse = inuse_bit_at_offset(nextchunk, nextsize);
3809 set_head(nextchunk, nextsize);
3810
3811 /* consolidate forward */
3812 if (!nextinuse) {
3813 unlink(nextchunk, bck, fwd);
3814 size += nextsize;
3815 }
3816
3817 /*
3818 Place the chunk in unsorted chunk list. Chunks are
3819 not placed into regular bins until after they have
3820 been given one chance to be used in malloc.
3821 */
3822
3823 bck = unsorted_chunks(av);
3824 fwd = bck->fd;
3825 p->bk = bck;
3826 p->fd = fwd;
3827 bck->fd = p;
3828 fwd->bk = p;
3829
3830 set_head(p, size | PREV_INUSE);
3831 set_foot(p, size);
3832
3833 check_free_chunk(p);
3834 }
3835
3836 /*
3837 If the chunk borders the current high end of memory,
3838 consolidate into top
3839 */
3840
3841 else {
3842 size += nextsize;
3843 set_head(p, size | PREV_INUSE);
3844 av->top = p;
3845 check_chunk(p);
3846 }
3847
3848 /*
3849 If freeing a large space, consolidate possibly-surrounding
3850 chunks. Then, if the total unused topmost memory exceeds trim
3851 threshold, ask malloc_trim to reduce top.
3852
3853 Unless max_fast is 0, we don't know if there are fastbins
3854 bordering top, so we cannot tell for sure whether threshold
3855 has been reached unless fastbins are consolidated. But we
3856 don't want to consolidate on each free. As a compromise,
3857 consolidation is performed if FASTBIN_CONSOLIDATION_THRESHOLD
3858 is reached.
3859 */
3860
3861 if ((CHUNK_SIZE_T)(size) >= FASTBIN_CONSOLIDATION_THRESHOLD) {
3862 if (have_fastchunks(av))
3863 malloc_consolidate(av);
3864
3865 #ifndef MORECORE_CANNOT_TRIM
3866 if ((CHUNK_SIZE_T)(chunksize(av->top)) >=
3867 (CHUNK_SIZE_T)(av->trim_threshold))
3868 sYSTRIm(av->top_pad, av);
3869 #endif
3870 }
3871
3872 }
3873 /*
3874 If the chunk was allocated via mmap, release via munmap()
3875 Note that if HAVE_MMAP is false but chunk_is_mmapped is
3876 true, then user must have overwritten memory. There's nothing
3877 we can do to catch this error unless DEBUG is set, in which case
3878 check_inuse_chunk (above) will have triggered error.
3879 */
3880
3881 else {
3882 #if HAVE_MMAP
3883 int ret;
3884 INTERNAL_SIZE_T offset = p->prev_size;
3885 av->n_mmaps--;
3886 av->mmapped_mem -= (size + offset);
3887 ret = munmap((char*)p - offset, size + offset);
3888 /* munmap returns non-zero on failure */
3889 assert(ret == 0);
3890 #endif
3891 }
3892 }
3893 }
3894
3895 /*
3896 ------------------------- malloc_consolidate -------------------------
3897
3898 malloc_consolidate is a specialized version of free() that tears
3899 down chunks held in fastbins. Free itself cannot be used for this
3900 purpose since, among other things, it might place chunks back onto
3901 fastbins. So, instead, we need to use a minor variant of the same
3902 code.
3903
3904 Also, because this routine needs to be called the first time through
3905 malloc anyway, it turns out to be the perfect place to trigger
3906 initialization code.
3907 */
3908
3909 #if __STD_C
3910 static void malloc_consolidate(mstate av)
3911 #else
3912 static void malloc_consolidate(av) mstate av;
3913 #endif
3914 {
3915 mfastbinptr* fb; /* current fastbin being consolidated */
3916 mfastbinptr* maxfb; /* last fastbin (for loop control) */
3917 mchunkptr p; /* current chunk being consolidated */
3918 mchunkptr nextp; /* next chunk to consolidate */
3919 mchunkptr unsorted_bin; /* bin header */
3920 mchunkptr first_unsorted; /* chunk to link to */
3921
3922 /* These have same use as in free() */
3923 mchunkptr nextchunk;
3924 INTERNAL_SIZE_T size;
3925 INTERNAL_SIZE_T nextsize;
3926 INTERNAL_SIZE_T prevsize;
3927 int nextinuse;
3928 mchunkptr bck;
3929 mchunkptr fwd;
3930
3931 /*
3932 If max_fast is 0, we know that av hasn't
3933 yet been initialized, in which case do so below
3934 */
3935
3936 if (av->max_fast != 0) {
3937 clear_fastchunks(av);
3938
3939 unsorted_bin = unsorted_chunks(av);
3940
3941 /*
3942 Remove each chunk from fast bin and consolidate it, placing it
3943 then in unsorted bin. Among other reasons for doing this,
3944 placing in unsorted bin avoids needing to calculate actual bins
3945 until malloc is sure that chunks aren't immediately going to be
3946 reused anyway.
3947 */
3948
3949 maxfb = &(av->fastbins[fastbin_index(av->max_fast)]);
3950 fb = &(av->fastbins[0]);
3951 do {
3952 if ( (p = *fb) != 0) {
3953 *fb = 0;
3954
3955 do {
3956 check_inuse_chunk(p);
3957 nextp = p->fd;
3958
3959 /* Slightly streamlined version of consolidation code in free() */
3960 size = p->size & ~PREV_INUSE;
3961 nextchunk = chunk_at_offset(p, size);
3962 nextsize = chunksize(nextchunk);
3963
3964 if (!prev_inuse(p)) {
3965 prevsize = p->prev_size;
3966 size += prevsize;
3967 p = chunk_at_offset(p, -((long) prevsize));
3968 unlink(p, bck, fwd);
3969 }
3970
3971 if (nextchunk != av->top) {
3972 nextinuse = inuse_bit_at_offset(nextchunk, nextsize);
3973 set_head(nextchunk, nextsize);
3974
3975 if (!nextinuse) {
3976 size += nextsize;
3977 unlink(nextchunk, bck, fwd);
3978 }
3979
3980 first_unsorted = unsorted_bin->fd;
3981 unsorted_bin->fd = p;
3982 first_unsorted->bk = p;
3983
3984 set_head(p, size | PREV_INUSE);
3985 p->bk = unsorted_bin;
3986 p->fd = first_unsorted;
3987 set_foot(p, size);
3988 }
3989
3990 else {
3991 size += nextsize;
3992 set_head(p, size | PREV_INUSE);
3993 av->top = p;
3994 }
3995
3996 } while ( (p = nextp) != 0);
3997
3998 }
3999 } while (fb++ != maxfb);
4000 }
4001 else {
4002 malloc_init_state(av);
4003 check_malloc_state();
4004 }
4005 }
4006
4007 /*
4008 ------------------------------ realloc ------------------------------
4009 */
4010
4011
4012 #if __STD_C
4013 Void_t* rEALLOc(Void_t* oldmem, size_t bytes)
4014 #else
4015 Void_t* rEALLOc(oldmem, bytes) Void_t* oldmem; size_t bytes;
4016 #endif
4017 {
4018 mstate av = get_malloc_state();
4019
4020 INTERNAL_SIZE_T nb; /* padded request size */
4021
4022 mchunkptr oldp; /* chunk corresponding to oldmem */
4023 INTERNAL_SIZE_T oldsize; /* its size */
4024
4025 mchunkptr newp; /* chunk to return */
4026 INTERNAL_SIZE_T newsize; /* its size */
4027 Void_t* newmem; /* corresponding user mem */
4028
4029 mchunkptr next; /* next contiguous chunk after oldp */
4030
4031 mchunkptr remainder; /* extra space at end of newp */
4032 CHUNK_SIZE_T remainder_size; /* its size */
4033
4034 mchunkptr bck; /* misc temp for linking */
4035 mchunkptr fwd; /* misc temp for linking */
4036
4037 CHUNK_SIZE_T copysize; /* bytes to copy */
4038 unsigned int ncopies; /* INTERNAL_SIZE_T words to copy */
4039 INTERNAL_SIZE_T* s; /* copy source */
4040 INTERNAL_SIZE_T* d; /* copy destination */
4041
4042 if (!__mpr_malloc_initialized)
4043 __mpr_malloc_initialize();
4044 if (__mpr_realloc_hook != NULL)
4045 return (*__mpr_realloc_hook)(oldmem, bytes);
4046
4047 #ifdef REALLOC_ZERO_BYTES_FREES
4048 if (bytes == 0) {
4049 fREe(oldmem);
4050 return 0;
4051 }
4052 #endif
4053
4054 /* realloc of null is supposed to be same as malloc */
4055 if (oldmem == 0) return mALLOc(bytes);
4056
4057 checked_request2size(bytes, nb);
4058
4059 oldp = mem2chunk(oldmem);
4060 oldsize = chunksize(oldp);
4061
4062 check_inuse_chunk(oldp);
4063
4064 if (!chunk_is_mmapped(oldp)) {
4065
4066 if ((CHUNK_SIZE_T)(oldsize) >= (CHUNK_SIZE_T)(nb)) {
4067 /* already big enough; split below */
4068 newp = oldp;
4069 newsize = oldsize;
4070 }
4071
4072 else {
4073 next = chunk_at_offset(oldp, oldsize);
4074
4075 /* Try to expand forward into top */
4076 if (next == av->top &&
4077 (CHUNK_SIZE_T)(newsize = oldsize + chunksize(next)) >=
4078 (CHUNK_SIZE_T)(nb + MINSIZE)) {
4079 set_head_size(oldp, nb);
4080 av->top = chunk_at_offset(oldp, nb);
4081 set_head(av->top, (newsize - nb) | PREV_INUSE);
4082 return chunk2mem(oldp);
4083 }
4084
4085 /* Try to expand forward into next chunk; split off remainder below */
4086 else if (next != av->top &&
4087 !inuse(next) &&
4088 (CHUNK_SIZE_T)(newsize = oldsize + chunksize(next)) >=
4089 (CHUNK_SIZE_T)(nb)) {
4090 newp = oldp;
4091 unlink(next, bck, fwd);
4092 }
4093
4094 /* allocate, copy, free */
4095 else {
4096 newmem = mALLOc(nb - MALLOC_ALIGN_MASK);
4097 if (newmem == 0)
4098 return 0; /* propagate failure */
4099
4100 newp = mem2chunk(newmem);
4101 newsize = chunksize(newp);
4102
4103 /*
4104 Avoid copy if newp is next chunk after oldp.
4105 */
4106 if (newp == next) {
4107 newsize += oldsize;
4108 newp = oldp;
4109 }
4110 else {
4111 /*
4112 Unroll copy of <= 36 bytes (72 if 8byte sizes)
4113 We know that contents have an odd number of
4114 INTERNAL_SIZE_T-sized words; minimally 3.
4115 */
4116
4117 copysize = oldsize - SIZE_SZ;
4118 s = (INTERNAL_SIZE_T*)(oldmem);
4119 d = (INTERNAL_SIZE_T*)(newmem);
4120 ncopies = copysize / sizeof(INTERNAL_SIZE_T);
4121 assert(ncopies >= 3);
4122
4123 if (ncopies > 9)
4124 MALLOC_COPY(d, s, copysize);
4125
4126 else {
4127 *(d+0) = *(s+0);
4128 *(d+1) = *(s+1);
4129 *(d+2) = *(s+2);
4130 if (ncopies > 4) {
4131 *(d+3) = *(s+3);
4132 *(d+4) = *(s+4);
4133 if (ncopies > 6) {
4134 *(d+5) = *(s+5);
4135 *(d+6) = *(s+6);
4136 if (ncopies > 8) {
4137 *(d+7) = *(s+7);
4138 *(d+8) = *(s+8);
4139 }
4140 }
4141 }
4142 }
4143
4144 fREe(oldmem);
4145 check_inuse_chunk(newp);
4146 return chunk2mem(newp);
4147 }
4148 }
4149 }
4150
4151 /* If possible, free extra space in old or extended chunk */
4152
4153 assert((CHUNK_SIZE_T)(newsize) >= (CHUNK_SIZE_T)(nb));
4154
4155 remainder_size = newsize - nb;
4156
4157 if (remainder_size < MINSIZE) { /* not enough extra to split off */
4158 set_head_size(newp, newsize);
4159 set_inuse_bit_at_offset(newp, newsize);
4160 }
4161 else { /* split remainder */
4162 remainder = chunk_at_offset(newp, nb);
4163 set_head_size(newp, nb);
4164 set_head(remainder, remainder_size | PREV_INUSE);
4165 /* Mark remainder as inuse so free() won't complain */
4166 set_inuse_bit_at_offset(remainder, remainder_size);
4167 fREe(chunk2mem(remainder));
4168 }
4169
4170 check_inuse_chunk(newp);
4171 return chunk2mem(newp);
4172 }
4173
4174 /*
4175 Handle mmap cases
4176 */
4177
4178 else {
4179 #if HAVE_MMAP
4180
4181 #if HAVE_MREMAP
4182 INTERNAL_SIZE_T offset = oldp->prev_size;
4183 size_t pagemask = av->pagesize - 1;
4184 char *cp;
4185 CHUNK_SIZE_T sum;
4186
4187 /* Note the extra SIZE_SZ overhead */
4188 newsize = (nb + offset + SIZE_SZ + pagemask) & ~pagemask;
4189
4190 /* don't need to remap if still within same page */
4191 if (oldsize == newsize - offset)
4192 return oldmem;
4193
4194 cp = (char*)mremap((char*)oldp - offset, oldsize + offset, newsize, 1);
4195
4196 if (cp != (char*)MORECORE_FAILURE) {
4197
4198 newp = (mchunkptr)(cp + offset);
4199 set_head(newp, (newsize - offset)|IS_MMAPPED);
4200
4201 assert(aligned_OK(chunk2mem(newp)));
4202 assert((newp->prev_size == offset));
4203
4204 /* update statistics */
4205 sum = av->mmapped_mem += newsize - oldsize;
4206 if (sum > (CHUNK_SIZE_T)(av->max_mmapped_mem))
4207 av->max_mmapped_mem = sum;
4208 sum += av->sbrked_mem;
4209 if (sum > (CHUNK_SIZE_T)(av->max_total_mem))
4210 av->max_total_mem = sum;
4211
4212 return chunk2mem(newp);
4213 }
4214 #endif
4215
4216 /* Note the extra SIZE_SZ overhead. */
4217 if ((CHUNK_SIZE_T)(oldsize) >= (CHUNK_SIZE_T)(nb + SIZE_SZ))
4218 newmem = oldmem; /* do nothing */
4219 else {
4220 /* Must alloc, copy, free. */
4221 newmem = mALLOc(nb - MALLOC_ALIGN_MASK);
4222 if (newmem != 0) {
4223 MALLOC_COPY(newmem, oldmem, oldsize - 2*SIZE_SZ);
4224 fREe(oldmem);
4225 }
4226 }
4227 return newmem;
4228
4229 #else
4230 /* If !HAVE_MMAP, but chunk_is_mmapped, user must have overwritten mem */
4231 check_malloc_state();
4232 MALLOC_FAILURE_ACTION;
4233 return 0;
4234 #endif
4235 }
4236 }
4237
4238 /*
4239 ------------------------------ memalign ------------------------------
4240 */
4241
4242 #if __STD_C
4243 Void_t* mEMALIGn(size_t alignment, size_t bytes)
4244 #else
4245 Void_t* mEMALIGn(alignment, bytes) size_t alignment; size_t bytes;
4246 #endif
4247 {
4248 INTERNAL_SIZE_T nb; /* padded request size */
4249 char* m; /* memory returned by malloc call */
4250 mchunkptr p; /* corresponding chunk */
4251 char* brk; /* alignment point within p */
4252 mchunkptr newp; /* chunk to return */
4253 INTERNAL_SIZE_T newsize; /* its size */
4254 INTERNAL_SIZE_T leadsize; /* leading space before alignment point */
4255 mchunkptr remainder; /* spare room at end to split off */
4256 CHUNK_SIZE_T remainder_size; /* its size */
4257 INTERNAL_SIZE_T size;
4258
4259 /* If need less alignment than we give anyway, just relay to malloc */
4260
4261 if (alignment <= MALLOC_ALIGNMENT) return mALLOc(bytes);
4262
4263 /* Otherwise, ensure that it is at least a minimum chunk size */
4264
4265 if (alignment < MINSIZE) alignment = MINSIZE;
4266
4267 /* Make sure alignment is power of 2 (in case MINSIZE is not). */
4268 if ((alignment & (alignment - 1)) != 0) {
4269 size_t a = MALLOC_ALIGNMENT * 2;
4270 while ((CHUNK_SIZE_T)a < (CHUNK_SIZE_T)alignment) a <<= 1;
4271 alignment = a;
4272 }
4273
4274 checked_request2size(bytes, nb);
4275
4276 /*
4277 Strategy: find a spot within that chunk that meets the alignment
4278 request, and then possibly free the leading and trailing space.
4279 */
4280
4281
4282 /* Call malloc with worst case padding to hit alignment. */
4283
4284 m = (char*)(mALLOc(nb + alignment + MINSIZE));
4285
4286 if (m == 0) return 0; /* propagate failure */
4287
4288 p = mem2chunk(m);
4289
4290 if ((((PTR_UINT)(m)) % alignment) != 0) { /* misaligned */
4291
4292 /*
4293 Find an aligned spot inside chunk. Since we need to give back
4294 leading space in a chunk of at least MINSIZE, if the first
4295 calculation places us at a spot with less than MINSIZE leader,
4296 we can move to the next aligned spot -- we've allocated enough
4297 total room so that this is always possible.
4298 */
4299
4300 brk = (char*)mem2chunk((PTR_UINT)(((PTR_UINT)(m + alignment - 1)) &
4301 -((signed long) alignment)));
4302 if ((CHUNK_SIZE_T)(brk - (char*)(p)) < MINSIZE)
4303 brk += alignment;
4304
4305 newp = (mchunkptr)brk;
4306 leadsize = brk - (char*)(p);
4307 newsize = chunksize(p) - leadsize;
4308
4309 /* For mmapped chunks, just adjust offset */
4310 if (chunk_is_mmapped(p)) {
4311 newp->prev_size = p->prev_size + leadsize;
4312 set_head(newp, newsize|IS_MMAPPED);
4313 return chunk2mem(newp);
4314 }
4315
4316 /* Otherwise, give back leader, use the rest */
4317 set_head(newp, newsize | PREV_INUSE);
4318 set_inuse_bit_at_offset(newp, newsize);
4319 set_head_size(p, leadsize);
4320 fREe(chunk2mem(p));
4321 p = newp;
4322
4323 assert (newsize >= nb &&
4324 (((PTR_UINT)(chunk2mem(p))) % alignment) == 0);
4325 }
4326
4327 /* Also give back spare room at the end */
4328 if (!chunk_is_mmapped(p)) {
4329 size = chunksize(p);
4330 if ((CHUNK_SIZE_T)(size) > (CHUNK_SIZE_T)(nb + MINSIZE)) {
4331 remainder_size = size - nb;
4332 remainder = chunk_at_offset(p, nb);
4333 set_head(remainder, remainder_size | PREV_INUSE);
4334 set_head_size(p, nb);
4335 fREe(chunk2mem(remainder));
4336 }
4337 }
4338
4339 check_inuse_chunk(p);
4340 return chunk2mem(p);
4341 }
4342
4343 /*
4344 ------------------------------ calloc ------------------------------
4345 */
4346
4347 #if __STD_C
4348 Void_t* cALLOc(size_t n_elements, size_t elem_size)
4349 #else
4350 Void_t* cALLOc(n_elements, elem_size) size_t n_elements; size_t elem_size;
4351 #endif
4352 {
4353 mchunkptr p;
4354 CHUNK_SIZE_T clearsize;
4355 CHUNK_SIZE_T nclears;
4356 INTERNAL_SIZE_T* d;
4357
4358 Void_t* mem = mALLOc(n_elements * elem_size);
4359
4360 if (mem != 0) {
4361 p = mem2chunk(mem);
4362
4363 if (!chunk_is_mmapped(p))
4364 {
4365 /*
4366 Unroll clear of <= 36 bytes (72 if 8byte sizes)
4367 We know that contents have an odd number of
4368 INTERNAL_SIZE_T-sized words; minimally 3.
4369 */
4370
4371 d = (INTERNAL_SIZE_T*)mem;
4372 clearsize = chunksize(p) - SIZE_SZ;
4373 nclears = clearsize / sizeof(INTERNAL_SIZE_T);
4374 assert(nclears >= 3);
4375
4376 if (nclears > 9)
4377 MALLOC_ZERO(d, clearsize);
4378
4379 else {
4380 *(d+0) = 0;
4381 *(d+1) = 0;
4382 *(d+2) = 0;
4383 if (nclears > 4) {
4384 *(d+3) = 0;
4385 *(d+4) = 0;
4386 if (nclears > 6) {
4387 *(d+5) = 0;
4388 *(d+6) = 0;
4389 if (nclears > 8) {
4390 *(d+7) = 0;
4391 *(d+8) = 0;
4392 }
4393 }
4394 }
4395 }
4396 }
4397 #if ! MMAP_CLEARS
4398 else
4399 {
4400 d = (INTERNAL_SIZE_T*)mem;
4401 /*
4402 Note the additional SIZE_SZ
4403 */
4404 clearsize = chunksize(p) - 2*SIZE_SZ;
4405 MALLOC_ZERO(d, clearsize);
4406 }
4407 #endif
4408 }
4409 return mem;
4410 }
4411
4412 /*
4413 ------------------------------ cfree ------------------------------
4414 */
4415
4416 #if __STD_C
4417 void cFREe(Void_t *mem)
4418 #else
4419 void cFREe(mem) Void_t *mem;
4420 #endif
4421 {
4422 fREe(mem);
4423 }
4424
4425 /*
4426 ------------------------- independent_calloc -------------------------
4427 */
4428
4429 #if __STD_C
4430 Void_t** iCALLOc(size_t n_elements, size_t elem_size, Void_t* chunks[])
4431 #else
4432 Void_t** iCALLOc(n_elements, elem_size, chunks) size_t n_elements; size_t elem_size; Void_t* chunks[];
4433 #endif
4434 {
4435 size_t sz = elem_size; /* serves as 1-element array */
4436 /* opts arg of 3 means all elements are same size, and should be cleared */
4437 return iALLOc(n_elements, &sz, 3, chunks);
4438 }
4439
4440 /*
4441 ------------------------- independent_comalloc -------------------------
4442 */
4443
4444 #if __STD_C
4445 Void_t** iCOMALLOc(size_t n_elements, size_t sizes[], Void_t* chunks[])
4446 #else
4447 Void_t** iCOMALLOc(n_elements, sizes, chunks) size_t n_elements; size_t sizes[]; Void_t* chunks[];
4448 #endif
4449 {
4450 return iALLOc(n_elements, sizes, 0, chunks);
4451 }
4452
4453
4454 /*
4455 ------------------------------ ialloc ------------------------------
4456 ialloc provides common support for independent_X routines, handling all of
4457 the combinations that can result.
4458
4459 The opts arg has:
4460 bit 0 set if all elements are same size (using sizes[0])
4461 bit 1 set if elements should be zeroed
4462 */
4463
4464
4465 #if __STD_C
4466 static Void_t** iALLOc(size_t n_elements,
4467 size_t* sizes,
4468 int opts,
4469 Void_t* chunks[])
4470 #else
4471 static Void_t** iALLOc(n_elements, sizes, opts, chunks) size_t n_elements; size_t* sizes; int opts; Void_t* chunks[];
4472 #endif
4473 {
4474 mstate av = get_malloc_state();
4475 INTERNAL_SIZE_T element_size; /* chunksize of each element, if all same */
4476 INTERNAL_SIZE_T contents_size; /* total size of elements */
4477 INTERNAL_SIZE_T array_size; /* request size of pointer array */
4478 Void_t* mem; /* malloced aggregate space */
4479 mchunkptr p; /* corresponding chunk */
4480 INTERNAL_SIZE_T remainder_size; /* remaining bytes while splitting */
4481 Void_t** marray; /* either "chunks" or malloced ptr array */
4482 mchunkptr array_chunk; /* chunk for malloced ptr array */
4483 int mmx; /* to disable mmap */
4484 INTERNAL_SIZE_T size;
4485 size_t i;
4486
4487 /* Ensure initialization */
4488 if (av->max_fast == 0) malloc_consolidate(av);
4489
4490 /* compute array length, if needed */
4491 if (chunks != 0) {
4492 if (n_elements == 0)
4493 return chunks; /* nothing to do */
4494 marray = chunks;
4495 array_size = 0;
4496 }
4497 else {
4498 /* if empty req, must still return chunk representing empty array */
4499 if (n_elements == 0)
4500 return (Void_t**) mALLOc(0);
4501 marray = 0;
4502 array_size = request2size(n_elements * (sizeof(Void_t*)));
4503 }
4504
4505 /* compute total element size */
4506 if (opts & 0x1) { /* all-same-size */
4507 element_size = request2size(*sizes);
4508 contents_size = n_elements * element_size;
4509 }
4510 else { /* add up all the sizes */
4511 element_size = 0;
4512 contents_size = 0;
4513 for (i = 0; i != n_elements; ++i)
4514 contents_size += request2size(sizes[i]);
4515 }
4516
4517 /* subtract out alignment bytes from total to minimize overallocation */
4518 size = contents_size + array_size - MALLOC_ALIGN_MASK;
4519
4520 /*
4521 Allocate the aggregate chunk.
4522 But first disable mmap so malloc won't use it, since
4523 we would not be able to later free/realloc space internal
4524 to a segregated mmap region.
4525 */
4526 mmx = av->n_mmaps_max; /* disable mmap */
4527 av->n_mmaps_max = 0;
4528 mem = mALLOc(size);
4529 av->n_mmaps_max = mmx; /* reset mmap */
4530 if (mem == 0)
4531 return 0;
4532
4533 p = mem2chunk(mem);
4534 assert(!chunk_is_mmapped(p));
4535 remainder_size = chunksize(p);
4536
4537 if (opts & 0x2) { /* optionally clear the elements */
4538 MALLOC_ZERO(mem, remainder_size - SIZE_SZ - array_size);
4539 }
4540
4541 /* If not provided, allocate the pointer array as final part of chunk */
4542 if (marray == 0) {
4543 array_chunk = chunk_at_offset(p, contents_size);
4544 marray = (Void_t**) (chunk2mem(array_chunk));
4545 set_head(array_chunk, (remainder_size - contents_size) | PREV_INUSE);
4546 remainder_size = contents_size;
4547 }
4548
4549 /* split out elements */
4550 for (i = 0; ; ++i) {
4551 marray[i] = chunk2mem(p);
4552 if (i != n_elements-1) {
4553 if (element_size != 0)
4554 size = element_size;
4555 else
4556 size = request2size(sizes[i]);
4557 remainder_size -= size;
4558 set_head(p, size | PREV_INUSE);
4559 p = chunk_at_offset(p, size);
4560 }
4561 else { /* the final element absorbs any overallocation slop */
4562 set_head(p, remainder_size | PREV_INUSE);
4563 break;
4564 }
4565 }
4566
4567 #if DEBUG
4568 if (marray != chunks) {
4569 /* final element must have exactly exhausted chunk */
4570 if (element_size != 0)
4571 assert(remainder_size == element_size);
4572 else
4573 assert(remainder_size == request2size(sizes[i]));
4574 check_inuse_chunk(mem2chunk(marray));
4575 }
4576
4577 for (i = 0; i != n_elements; ++i)
4578 check_inuse_chunk(mem2chunk(marray[i]));
4579 #endif
4580
4581 return marray;
4582 }
4583
4584
4585 /*
4586 ------------------------------ valloc ------------------------------
4587 */
4588
4589 #if __STD_C
4590 Void_t* vALLOc(size_t bytes)
4591 #else
4592 Void_t* vALLOc(bytes) size_t bytes;
4593 #endif
4594 {
4595 /* Ensure initialization */
4596 mstate av = get_malloc_state();
4597 if (av->max_fast == 0) malloc_consolidate(av);
4598 return mEMALIGn(av->pagesize, bytes);
4599 }
4600
4601 /*
4602 ------------------------------ pvalloc ------------------------------
4603 */
4604
4605
4606 #if __STD_C
4607 Void_t* pVALLOc(size_t bytes)
4608 #else
4609 Void_t* pVALLOc(bytes) size_t bytes;
4610 #endif
4611 {
4612 mstate av = get_malloc_state();
4613 size_t pagesz;
4614
4615 /* Ensure initialization */
4616 if (av->max_fast == 0) malloc_consolidate(av);
4617 pagesz = av->pagesize;
4618 return mEMALIGn(pagesz, (bytes + pagesz - 1) & ~(pagesz - 1));
4619 }
4620
4621
4622 /*
4623 ------------------------------ malloc_trim ------------------------------
4624 */
4625
4626 #if __STD_C
4627 int mTRIm(size_t pad)
4628 #else
4629 int mTRIm(pad) size_t pad;
4630 #endif
4631 {
4632 mstate av = get_malloc_state();
4633 /* Ensure initialization/consolidation */
4634 malloc_consolidate(av);
4635
4636 #ifndef MORECORE_CANNOT_TRIM
4637 return sYSTRIm(pad, av);
4638 #else
4639 return 0;
4640 #endif
4641 }
4642
4643
4644 /*
4645 ------------------------- malloc_usable_size -------------------------
4646 */
4647
4648 #if __STD_C
4649 size_t mUSABLe(Void_t* mem)
4650 #else
4651 size_t mUSABLe(mem) Void_t* mem;
4652 #endif
4653 {
4654 mchunkptr p;
4655 if (mem != 0) {
4656 p = mem2chunk(mem);
4657 if (chunk_is_mmapped(p))
4658 return chunksize(p) - 2*SIZE_SZ;
4659 else if (inuse(p))
4660 return chunksize(p) - SIZE_SZ;
4661 }
4662 return 0;
4663 }
4664
4665 /*
4666 ------------------------------ mallinfo ------------------------------
4667 */
4668
4669 struct mallinfo mALLINFo()
4670 {
4671 mstate av = get_malloc_state();
4672 struct mallinfo mi;
4673 int i;
4674 mbinptr b;
4675 mchunkptr p;
4676 INTERNAL_SIZE_T avail;
4677 INTERNAL_SIZE_T fastavail;
4678 int nblocks;
4679 int nfastblocks;
4680
4681 /* Ensure initialization */
4682 if (av->top == 0) malloc_consolidate(av);
4683
4684 check_malloc_state();
4685
4686 /* Account for top */
4687 avail = chunksize(av->top);
4688 nblocks = 1; /* top always exists */
4689
4690 /* traverse fastbins */
4691 nfastblocks = 0;
4692 fastavail = 0;
4693
4694 for (i = 0; i < NFASTBINS; ++i) {
4695 for (p = av->fastbins[i]; p != 0; p = p->fd) {
4696 ++nfastblocks;
4697 fastavail += chunksize(p);
4698 }
4699 }
4700
4701 avail += fastavail;
4702
4703 /* traverse regular bins */
4704 for (i = 1; i < NBINS; ++i) {
4705 b = bin_at(av, i);
4706 for (p = last(b); p != b; p = p->bk) {
4707 ++nblocks;
4708 avail += chunksize(p);
4709 }
4710 }
4711
4712 mi.smblks = nfastblocks;
4713 mi.ordblks = nblocks;
4714 mi.fordblks = avail;
4715 mi.uordblks = av->sbrked_mem - avail;
4716 mi.arena = av->sbrked_mem;
4717 mi.hblks = av->n_mmaps;
4718 mi.hblkhd = av->mmapped_mem;
4719 mi.fsmblks = fastavail;
4720 mi.keepcost = chunksize(av->top);
4721 mi.usmblks = av->max_total_mem;
4722 return mi;
4723 }
4724
4725 /*
4726 ------------------------------ malloc_stats ------------------------------
4727 */
4728
4729 void mSTATs()
4730 {
4731 struct mallinfo mi = mALLINFo();
4732
4733 #ifdef WIN32
4734 {
4735 CHUNK_SIZE_T free, reserved, committed;
4736 vminfo (&free, &reserved, &committed);
4737 fprintf(stderr, "free bytes = %10lu\n",
4738 free);
4739 fprintf(stderr, "reserved bytes = %10lu\n",
4740 reserved);
4741 fprintf(stderr, "committed bytes = %10lu\n",
4742 committed);
4743 }
4744 #endif
4745
4746
4747 fprintf(stderr, "max system bytes = %10lu\n",
4748 (CHUNK_SIZE_T)(mi.usmblks));
4749 fprintf(stderr, "system bytes = %10lu\n",
4750 (CHUNK_SIZE_T)(mi.arena + mi.hblkhd));
4751 fprintf(stderr, "in use bytes = %10lu\n",
4752 (CHUNK_SIZE_T)(mi.uordblks + mi.hblkhd));
4753
4754 #ifdef WIN32
4755 {
4756 CHUNK_SIZE_T kernel, user;
4757 if (cpuinfo (TRUE, &kernel, &user)) {
4758 fprintf(stderr, "kernel ms = %10lu\n",
4759 kernel);
4760 fprintf(stderr, "user ms = %10lu\n",
4761 user);
4762 }
4763 }
4764 #endif
4765 }
4766
4767
4768 /*
4769 ------------------------------ mallopt ------------------------------
4770 */
4771
4772 #if __STD_C
4773 int mALLOPt(int param_number, int value)
4774 #else
4775 int mALLOPt(param_number, value) int param_number; int value;
4776 #endif
4777 {
4778 mstate av = get_malloc_state();
4779 /* Ensure initialization/consolidation */
4780 malloc_consolidate(av);
4781
4782 switch(param_number) {
4783 case M_MXFAST:
4784 if (value >= 0 && value <= MAX_FAST_SIZE) {
4785 set_max_fast(av, value);
4786 return 1;
4787 }
4788 else
4789 return 0;
4790
4791 case M_TRIM_THRESHOLD:
4792 av->trim_threshold = value;
4793 return 1;
4794
4795 case M_TOP_PAD:
4796 av->top_pad = value;
4797 return 1;
4798
4799 case M_MMAP_THRESHOLD:
4800 av->mmap_threshold = value;
4801 return 1;
4802
4803 case M_MMAP_MAX:
4804 #if !HAVE_MMAP
4805 if (value != 0)
4806 return 0;
4807 #endif
4808 av->n_mmaps_max = value;
4809 return 1;
4810
4811 default:
4812 return 0;
4813 }
4814 }
4815
4816
4817 /*
4818 -------------------- Alternative MORECORE functions --------------------
4819 */
4820
4821
4822 /*
4823 General Requirements for MORECORE.
4824
4825 The MORECORE function must have the following properties:
4826
4827 If MORECORE_CONTIGUOUS is false:
4828
4829 * MORECORE must allocate in multiples of pagesize. It will
4830 only be called with arguments that are multiples of pagesize.
4831
4832 * MORECORE(0) must return an address that is at least
4833 MALLOC_ALIGNMENT aligned. (Page-aligning always suffices.)
4834
4835 else (i.e. If MORECORE_CONTIGUOUS is true):
4836
4837 * Consecutive calls to MORECORE with positive arguments
4838 return increasing addresses, indicating that space has been
4839 contiguously extended.
4840
4841 * MORECORE need not allocate in multiples of pagesize.
4842 Calls to MORECORE need not have args of multiples of pagesize.
4843
4844 * MORECORE need not page-align.
4845
4846 In either case:
4847
4848 * MORECORE may allocate more memory than requested. (Or even less,
4849 but this will generally result in a malloc failure.)
4850
4851 * MORECORE must not allocate memory when given argument zero, but
4852 instead return one past the end address of memory from previous
4853 nonzero call. This malloc does NOT call MORECORE(0)
4854 until at least one call with positive arguments is made, so
4855 the initial value returned is not important.
4856
4857 * Even though consecutive calls to MORECORE need not return contiguous
4858 addresses, it must be OK for malloc'ed chunks to span multiple
4859 regions in those cases where they do happen to be contiguous.
4860
4861 * MORECORE need not handle negative arguments -- it may instead
4862 just return MORECORE_FAILURE when given negative arguments.
4863 Negative arguments are always multiples of pagesize. MORECORE
4864 must not misinterpret negative args as large positive unsigned
4865 args. You can suppress all such calls from even occurring by defining
4866 MORECORE_CANNOT_TRIM,
4867
4868 There is some variation across systems about the type of the
4869 argument to sbrk/MORECORE. If size_t is unsigned, then it cannot
4870 actually be size_t, because sbrk supports negative args, so it is
4871 normally the signed type of the same width as size_t (sometimes
4872 declared as "intptr_t", and sometimes "ptrdiff_t"). It doesn't much
4873 matter though. Internally, we use "long" as arguments, which should
4874 work across all reasonable possibilities.
4875
4876 Additionally, if MORECORE ever returns failure for a positive
4877 request, and HAVE_MMAP is true, then mmap is used as a noncontiguous
4878 system allocator. This is a useful backup strategy for systems with
4879 holes in address spaces -- in this case sbrk cannot contiguously
4880 expand the heap, but mmap may be able to map noncontiguous space.
4881
4882 If you'd like mmap to ALWAYS be used, you can define MORECORE to be
4883 a function that always returns MORECORE_FAILURE.
4884
4885 Malloc only has limited ability to detect failures of MORECORE
4886 to supply contiguous space when it says it can. In particular,
4887 multithreaded programs that do not use locks may result in
4888 rece conditions across calls to MORECORE that result in gaps
4889 that cannot be detected as such, and subsequent corruption.
4890
4891 If you are using this malloc with something other than sbrk (or its
4892 emulation) to supply memory regions, you probably want to set
4893 MORECORE_CONTIGUOUS as false. As an example, here is a custom
4894 allocator kindly contributed for pre-OSX macOS. It uses virtually
4895 but not necessarily physically contiguous non-paged memory (locked
4896 in, present and won't get swapped out). You can use it by
4897 uncommenting this section, adding some #includes, and setting up the
4898 appropriate defines above:
4899
4900 #define MORECORE osMoreCore
4901 #define MORECORE_CONTIGUOUS 0
4902
4903 There is also a shutdown routine that should somehow be called for
4904 cleanup upon program exit.
4905
4906 #define MAX_POOL_ENTRIES 100
4907 #define MINIMUM_MORECORE_SIZE (64 * 1024)
4908 static int next_os_pool;
4909 void *our_os_pools[MAX_POOL_ENTRIES];
4910
4911 void *osMoreCore(int size)
4912 {
4913 void *ptr = 0;
4914 static void *sbrk_top = 0;
4915
4916 if (size > 0)
4917 {
4918 if (size < MINIMUM_MORECORE_SIZE)
4919 size = MINIMUM_MORECORE_SIZE;
4920 if (CurrentExecutionLevel() == kTaskLevel)
4921 ptr = PoolAllocateResident(size + RM_PAGE_SIZE, 0);
4922 if (ptr == 0)
4923 {
4924 return (void *) MORECORE_FAILURE;
4925 }
4926 // save ptrs so they can be freed during cleanup
4927 our_os_pools[next_os_pool] = ptr;
4928 next_os_pool++;
4929 ptr = (void *) ((((CHUNK_SIZE_T) ptr) + RM_PAGE_MASK) & ~RM_PAGE_MASK);
4930 sbrk_top = (char *) ptr + size;
4931 return ptr;
4932 }
4933 else if (size < 0)
4934 {
4935 // we don't currently support shrink behavior
4936 return (void *) MORECORE_FAILURE;
4937 }
4938 else
4939 {
4940 return sbrk_top;
4941 }
4942 }
4943
4944 // cleanup any allocated memory pools
4945 // called as last thing before shutting down driver
4946
4947 void osCleanupMem(void)
4948 {
4949 void **ptr;
4950
4951 for (ptr = our_os_pools; ptr < &our_os_pools[MAX_POOL_ENTRIES]; ptr++)
4952 if (*ptr)
4953 {
4954 PoolDeallocate(*ptr);
4955 *ptr = 0;
4956 }
4957 }
4958
4959 */
4960
4961
4962 /*
4963 --------------------------------------------------------------
4964
4965 Emulation of sbrk for win32.
4966 Donated by J. Walter <Walter@GeNeSys-e.de>.
4967 For additional information about this code, and malloc on Win32, see
4968 http://www.genesys-e.de/jwalter/
4969 */
4970
4971
4972 #ifdef WIN32
4973
4974 #ifdef _DEBUG
4975 /* #define TRACE */
4976 #endif
4977
4978 /* Support for USE_MALLOC_LOCK */
4979 #ifdef USE_MALLOC_LOCK
4980
4981 /* Wait for spin lock */
4982 static int slwait (int *sl) {
4983 while (InterlockedCompareExchange ((void **) sl, (void *) 1, (void *) 0) != 0)
4984 Sleep (0);
4985 return 0;
4986 }
4987
4988 /* Release spin lock */
4989 static int slrelease (int *sl) {
4990 InterlockedExchange (sl, 0);
4991 return 0;
4992 }
4993
4994 #ifdef NEEDED
4995 /* Spin lock for emulation code */
4996 static int g_sl;
4997 #endif
4998
4999 #endif /* USE_MALLOC_LOCK */
5000
5001 /* getpagesize for windows */
5002 static long getpagesize (void) {
5003 static long g_pagesize = 0;
5004 if (! g_pagesize) {
5005 SYSTEM_INFO system_info;
5006 GetSystemInfo (&system_info);
5007 g_pagesize = system_info.dwPageSize;
5008 }
5009 return g_pagesize;
5010 }
5011 static long getregionsize (void) {
5012 static long g_regionsize = 0;
5013 if (! g_regionsize) {
5014 SYSTEM_INFO system_info;
5015 GetSystemInfo (&system_info);
5016 g_regionsize = system_info.dwAllocationGranularity;
5017 }
5018 return g_regionsize;
5019 }
5020
5021 /* A region list entry */
5022 typedef struct _region_list_entry {
5023 void *top_allocated;
5024 void *top_committed;
5025 void *top_reserved;
5026 long reserve_size;
5027 struct _region_list_entry *previous;
5028 } region_list_entry;
5029
5030 /* Allocate and link a region entry in the region list */
5031 static int region_list_append (region_list_entry **last, void *base_reserved, long reserve_size) {
5032 region_list_entry *next = HeapAlloc (GetProcessHeap (), 0, sizeof (region_list_entry));
5033 if (! next)
5034 return FALSE;
5035 next->top_allocated = (char *) base_reserved;
5036 next->top_committed = (char *) base_reserved;
5037 next->top_reserved = (char *) base_reserved + reserve_size;
5038 next->reserve_size = reserve_size;
5039 next->previous = *last;
5040 *last = next;
5041 return TRUE;
5042 }
5043 /* Free and unlink the last region entry from the region list */
5044 static int region_list_remove (region_list_entry **last) {
5045 region_list_entry *previous = (*last)->previous;
5046 if (! HeapFree (GetProcessHeap (), sizeof (region_list_entry), *last))
5047 return FALSE;
5048 *last = previous;
5049 return TRUE;
5050 }
5051
5052 #define CEIL(size,to) (((size)+(to)-1)&~((to)-1))
5053 #define FLOOR(size,to) ((size)&~((to)-1))
5054
5055 #define SBRK_SCALE 0
5056 /* #define SBRK_SCALE 1 */
5057 /* #define SBRK_SCALE 2 */
5058 /* #define SBRK_SCALE 4 */
5059
5060 /* sbrk for windows */
5061 static void *sbrk (long size) {
5062 static long g_pagesize, g_my_pagesize;
5063 static long g_regionsize, g_my_regionsize;
5064 static region_list_entry *g_last;
5065 void *result = (void *) MORECORE_FAILURE;
5066 #ifdef TRACE
5067 printf ("sbrk %d\n", size);
5068 #endif
5069 #if defined (USE_MALLOC_LOCK) && defined (NEEDED)
5070 /* Wait for spin lock */
5071 slwait (&g_sl);
5072 #endif
5073 /* First time initialization */
5074 if (! g_pagesize) {
5075 g_pagesize = getpagesize ();
5076 g_my_pagesize = g_pagesize << SBRK_SCALE;
5077 }
5078 if (! g_regionsize) {
5079 g_regionsize = getregionsize ();
5080 g_my_regionsize = g_regionsize << SBRK_SCALE;
5081 }
5082 if (! g_last) {
5083 if (! region_list_append (&g_last, 0, 0))
5084 goto sbrk_exit;
5085 }
5086 /* Assert invariants */
5087 assert (g_last);
5088 assert ((char *) g_last->top_reserved - g_last->reserve_size <= (char *) g_last->top_allocated &&
5089 g_last->top_allocated <= g_last->top_committed);
5090 assert ((char *) g_last->top_reserved - g_last->reserve_size <= (char *) g_last->top_committed &&
5091 g_last->top_committed <= g_last->top_reserved &&
5092 (unsigned) g_last->top_committed % g_pagesize == 0);
5093 assert ((unsigned) g_last->top_reserved % g_regionsize == 0);
5094 assert ((unsigned) g_last->reserve_size % g_regionsize == 0);
5095 /* Allocation requested? */
5096 if (size >= 0) {
5097 /* Allocation size is the requested size */
5098 long allocate_size = size;
5099 /* Compute the size to commit */
5100 long to_commit = (char *) g_last->top_allocated + allocate_size - (char *) g_last->top_committed;
5101 /* Do we reach the commit limit? */
5102 if (to_commit > 0) {
5103 /* Round size to commit */
5104 long commit_size = CEIL (to_commit, g_my_pagesize);
5105 /* Compute the size to reserve */
5106 long to_reserve = (char *) g_last->top_committed + commit_size - (char *) g_last->top_reserved;
5107 /* Do we reach the reserve limit? */
5108 if (to_reserve > 0) {
5109 /* Compute the remaining size to commit in the current region */
5110 long remaining_commit_size = (char *) g_last->top_reserved - (char *) g_last->top_committed;
5111 if (remaining_commit_size > 0) {
5112 /* Assert preconditions */
5113 assert ((unsigned) g_last->top_committed % g_pagesize == 0);
5114 assert (0 < remaining_commit_size && remaining_commit_size % g_pagesize == 0); {
5115 /* Commit this */
5116 void *base_committed = VirtualAlloc (g_last->top_committed, remaining_commit_size,
5117 MEM_COMMIT, PAGE_READWRITE);
5118 /* Check returned pointer for consistency */
5119 if (base_committed != g_last->top_committed)
5120 goto sbrk_exit;
5121 /* Assert postconditions */
5122 assert ((unsigned) base_committed % g_pagesize == 0);
5123 #ifdef TRACE
5124 printf ("Commit %p %d\n", base_committed, remaining_commit_size);
5125 #endif
5126 /* Adjust the regions commit top */
5127 g_last->top_committed = (char *) base_committed + remaining_commit_size;
5128 }
5129 } {
5130 /* Now we are going to search and reserve. */
5131 int contiguous = -1;
5132 int found = FALSE;
5133 MEMORY_BASIC_INFORMATION memory_info;
5134 void *base_reserved;
5135 long reserve_size;
5136 do {
5137 /* Assume contiguous memory */
5138 contiguous = TRUE;
5139 /* Round size to reserve */
5140 reserve_size = CEIL (to_reserve, g_my_regionsize);
5141 /* Start with the current region's top */
5142 memory_info.BaseAddress = g_last->top_reserved;
5143 /* Assert preconditions */
5144 assert ((unsigned) memory_info.BaseAddress % g_pagesize == 0);
5145 assert (0 < reserve_size && reserve_size % g_regionsize == 0);
5146 while (VirtualQuery (memory_info.BaseAddress, &memory_info, sizeof (memory_info))) {
5147 /* Assert postconditions */
5148 assert ((unsigned) memory_info.BaseAddress % g_pagesize == 0);
5149 #ifdef TRACE
5150 printf ("Query %p %d %s\n", memory_info.BaseAddress, memory_info.RegionSize,
5151 memory_info.State == MEM_FREE ? "FREE":
5152 (memory_info.State == MEM_RESERVE ? "RESERVED":
5153 (memory_info.State == MEM_COMMIT ? "COMMITTED": "?")));
5154 #endif
5155 /* Region is free, well aligned and big enough: we are done */
5156 if (memory_info.State == MEM_FREE &&
5157 (unsigned) memory_info.BaseAddress % g_regionsize == 0 &&
5158 memory_info.RegionSize >= (unsigned) reserve_size) {
5159 found = TRUE;
5160 break;
5161 }
5162 /* From now on we can't get contiguous memory! */
5163 contiguous = FALSE;
5164 /* Recompute size to reserve */
5165 reserve_size = CEIL (allocate_size, g_my_regionsize);
5166 memory_info.BaseAddress = (char *) memory_info.BaseAddress + memory_info.RegionSize;
5167 /* Assert preconditions */
5168 assert ((unsigned) memory_info.BaseAddress % g_pagesize == 0);
5169 assert (0 < reserve_size && reserve_size % g_regionsize == 0);
5170 }
5171 /* Search failed? */
5172 if (! found)
5173 goto sbrk_exit;
5174 /* Assert preconditions */
5175 assert ((unsigned) memory_info.BaseAddress % g_regionsize == 0);
5176 assert (0 < reserve_size && reserve_size % g_regionsize == 0);
5177 /* Try to reserve this */
5178 base_reserved = VirtualAlloc (memory_info.BaseAddress, reserve_size,
5179 MEM_RESERVE, PAGE_NOACCESS);
5180 if (! base_reserved) {
5181 int rc = GetLastError ();
5182 if (rc != ERROR_INVALID_ADDRESS)
5183 goto sbrk_exit;
5184 }
5185 /* A null pointer signals (hopefully) a race condition with another thread. */
5186 /* In this case, we try again. */
5187 } while (! base_reserved);
5188 /* Check returned pointer for consistency */
5189 if (memory_info.BaseAddress && base_reserved != memory_info.BaseAddress)
5190 goto sbrk_exit;
5191 /* Assert postconditions */
5192 assert ((unsigned) base_reserved % g_regionsize == 0);
5193 #ifdef TRACE
5194 printf ("Reserve %p %d\n", base_reserved, reserve_size);
5195 #endif
5196 /* Did we get contiguous memory? */
5197 if (contiguous) {
5198 long start_size = (char *) g_last->top_committed - (char *) g_last->top_allocated;
5199 /* Adjust allocation size */
5200 allocate_size -= start_size;
5201 /* Adjust the regions allocation top */
5202 g_last->top_allocated = g_last->top_committed;
5203 /* Recompute the size to commit */
5204 to_commit = (char *) g_last->top_allocated + allocate_size - (char *) g_last->top_committed;
5205 /* Round size to commit */
5206 commit_size = CEIL (to_commit, g_my_pagesize);
5207 }
5208 /* Append the new region to the list */
5209 if (! region_list_append (&g_last, base_reserved, reserve_size))
5210 goto sbrk_exit;
5211 /* Didn't we get contiguous memory? */
5212 if (! contiguous) {
5213 /* Recompute the size to commit */
5214 to_commit = (char *) g_last->top_allocated + allocate_size - (char *) g_last->top_committed;
5215 /* Round size to commit */
5216 commit_size = CEIL (to_commit, g_my_pagesize);
5217 }
5218 }
5219 }
5220 /* Assert preconditions */
5221 assert ((unsigned) g_last->top_committed % g_pagesize == 0);
5222 assert (0 < commit_size && commit_size % g_pagesize == 0); {
5223 /* Commit this */
5224 void *base_committed = VirtualAlloc (g_last->top_committed, commit_size,
5225 MEM_COMMIT, PAGE_READWRITE);
5226 /* Check returned pointer for consistency */
5227 if (base_committed != g_last->top_committed)
5228 goto sbrk_exit;
5229 /* Assert postconditions */
5230 assert ((unsigned) base_committed % g_pagesize == 0);
5231 #ifdef TRACE
5232 printf ("Commit %p %d\n", base_committed, commit_size);
5233 #endif
5234 /* Adjust the regions commit top */
5235 g_last->top_committed = (char *) base_committed + commit_size;
5236 }
5237 }
5238 /* Adjust the regions allocation top */
5239 g_last->top_allocated = (char *) g_last->top_allocated + allocate_size;
5240 result = (char *) g_last->top_allocated - size;
5241 /* Deallocation requested? */
5242 } else if (size < 0) {
5243 long deallocate_size = - size;
5244 /* As long as we have a region to release */
5245 while ((char *) g_last->top_allocated - deallocate_size < (char *) g_last->top_reserved - g_last->reserve_size) {
5246 /* Get the size to release */
5247 long release_size = g_last->reserve_size;
5248 /* Get the base address */
5249 void *base_reserved = (char *) g_last->top_reserved - release_size;
5250 /* Assert preconditions */
5251 assert ((unsigned) base_reserved % g_regionsize == 0);
5252 assert (0 < release_size && release_size % g_regionsize == 0); {
5253 /* Release this */
5254 int rc = VirtualFree (base_reserved, 0,
5255 MEM_RELEASE);
5256 /* Check returned code for consistency */
5257 if (! rc)
5258 goto sbrk_exit;
5259 #ifdef TRACE
5260 printf ("Release %p %d\n", base_reserved, release_size);
5261 #endif
5262 }
5263 /* Adjust deallocation size */
5264 deallocate_size -= (char *) g_last->top_allocated - (char *) base_reserved;
5265 /* Remove the old region from the list */
5266 if (! region_list_remove (&g_last))
5267 goto sbrk_exit;
5268 } {
5269 /* Compute the size to decommit */
5270 long to_decommit = (char *) g_last->top_committed - ((char *) g_last->top_allocated - deallocate_size);
5271 if (to_decommit >= g_my_pagesize) {
5272 /* Compute the size to decommit */
5273 long decommit_size = FLOOR (to_decommit, g_my_pagesize);
5274 /* Compute the base address */
5275 void *base_committed = (char *) g_last->top_committed - decommit_size;
5276 /* Assert preconditions */
5277 assert ((unsigned) base_committed % g_pagesize == 0);
5278 assert (0 < decommit_size && decommit_size % g_pagesize == 0); {
5279 /* Decommit this */
5280 int rc = VirtualFree ((char *) base_committed, decommit_size,
5281 MEM_DECOMMIT);
5282 /* Check returned code for consistency */
5283 if (! rc)
5284 goto sbrk_exit;
5285 #ifdef TRACE
5286 printf ("Decommit %p %d\n", base_committed, decommit_size);
5287 #endif
5288 }
5289 /* Adjust deallocation size and regions commit and allocate top */
5290 deallocate_size -= (char *) g_last->top_allocated - (char *) base_committed;
5291 g_last->top_committed = base_committed;
5292 g_last->top_allocated = base_committed;
5293 }
5294 }
5295 /* Adjust regions allocate top */
5296 g_last->top_allocated = (char *) g_last->top_allocated - deallocate_size;
5297 /* Check for underflow */
5298 if ((char *) g_last->top_reserved - g_last->reserve_size > (char *) g_last->top_allocated ||
5299 g_last->top_allocated > g_last->top_committed) {
5300 /* Adjust regions allocate top */
5301 g_last->top_allocated = (char *) g_last->top_reserved - g_last->reserve_size;
5302 goto sbrk_exit;
5303 }
5304 result = g_last->top_allocated;
5305 }
5306 /* Assert invariants */
5307 assert (g_last);
5308 assert ((char *) g_last->top_reserved - g_last->reserve_size <= (char *) g_last->top_allocated &&
5309 g_last->top_allocated <= g_last->top_committed);
5310 assert ((char *) g_last->top_reserved - g_last->reserve_size <= (char *) g_last->top_committed &&
5311 g_last->top_committed <= g_last->top_reserved &&
5312 (unsigned) g_last->top_committed % g_pagesize == 0);
5313 assert ((unsigned) g_last->top_reserved % g_regionsize == 0);
5314 assert ((unsigned) g_last->reserve_size % g_regionsize == 0);
5315
5316 sbrk_exit:
5317 #if defined (USE_MALLOC_LOCK) && defined (NEEDED)
5318 /* Release spin lock */
5319 slrelease (&g_sl);
5320 #endif
5321 return result;
5322 }
5323
5324 /* mmap for windows */
5325 static void *mmap (void *ptr, long size, long prot, long type, long handle, long arg) {
5326 static long g_pagesize;
5327 static long g_regionsize;
5328 #ifdef TRACE
5329 printf ("mmap %d\n", size);
5330 #endif
5331 #if defined (USE_MALLOC_LOCK) && defined (NEEDED)
5332 /* Wait for spin lock */
5333 slwait (&g_sl);
5334 #endif
5335 /* First time initialization */
5336 if (! g_pagesize)
5337 g_pagesize = getpagesize ();
5338 if (! g_regionsize)
5339 g_regionsize = getregionsize ();
5340 /* Assert preconditions */
5341 assert ((unsigned) ptr % g_regionsize == 0);
5342 assert (size % g_pagesize == 0);
5343 /* Allocate this */
5344 ptr = VirtualAlloc (ptr, size,
5345 MEM_RESERVE | MEM_COMMIT | MEM_TOP_DOWN, PAGE_READWRITE);
5346 if (! ptr) {
5347 ptr = (void *) MORECORE_FAILURE;
5348 goto mmap_exit;
5349 }
5350 /* Assert postconditions */
5351 assert ((unsigned) ptr % g_regionsize == 0);
5352 #ifdef TRACE
5353 printf ("Commit %p %d\n", ptr, size);
5354 #endif
5355 mmap_exit:
5356 #if defined (USE_MALLOC_LOCK) && defined (NEEDED)
5357 /* Release spin lock */
5358 slrelease (&g_sl);
5359 #endif
5360 return ptr;
5361 }
5362
5363 /* munmap for windows */
5364 static long munmap (void *ptr, long size) {
5365 static long g_pagesize;
5366 static long g_regionsize;
5367 int rc = MUNMAP_FAILURE;
5368 #ifdef TRACE
5369 printf ("munmap %p %d\n", ptr, size);
5370 #endif
5371 #if defined (USE_MALLOC_LOCK) && defined (NEEDED)
5372 /* Wait for spin lock */
5373 slwait (&g_sl);
5374 #endif
5375 /* First time initialization */
5376 if (! g_pagesize)
5377 g_pagesize = getpagesize ();
5378 if (! g_regionsize)
5379 g_regionsize = getregionsize ();
5380 /* Assert preconditions */
5381 assert ((unsigned) ptr % g_regionsize == 0);
5382 assert (size % g_pagesize == 0);
5383 /* Free this */
5384 if (! VirtualFree (ptr, 0,
5385 MEM_RELEASE))
5386 goto munmap_exit;
5387 rc = 0;
5388 #ifdef TRACE
5389 printf ("Release %p %d\n", ptr, size);
5390 #endif
5391 munmap_exit:
5392 #if defined (USE_MALLOC_LOCK) && defined (NEEDED)
5393 /* Release spin lock */
5394 slrelease (&g_sl);
5395 #endif
5396 return rc;
5397 }
5398
5399 static void vminfo (CHUNK_SIZE_T *free, CHUNK_SIZE_T *reserved, CHUNK_SIZE_T *committed) {
5400 MEMORY_BASIC_INFORMATION memory_info;
5401 memory_info.BaseAddress = 0;
5402 *free = *reserved = *committed = 0;
5403 while (VirtualQuery (memory_info.BaseAddress, &memory_info, sizeof (memory_info))) {
5404 switch (memory_info.State) {
5405 case MEM_FREE:
5406 *free += memory_info.RegionSize;
5407 break;
5408 case MEM_RESERVE:
5409 *reserved += memory_info.RegionSize;
5410 break;
5411 case MEM_COMMIT:
5412 *committed += memory_info.RegionSize;
5413 break;
5414 }
5415 memory_info.BaseAddress = (char *) memory_info.BaseAddress + memory_info.RegionSize;
5416 }
5417 }
5418
5419 static int cpuinfo (int whole, CHUNK_SIZE_T *kernel, CHUNK_SIZE_T *user) {
5420 if (whole) {
5421 __int64 creation64, exit64, kernel64, user64;
5422 int rc = GetProcessTimes (GetCurrentProcess (),
5423 (FILETIME *) &creation64,
5424 (FILETIME *) &exit64,
5425 (FILETIME *) &kernel64,
5426 (FILETIME *) &user64);
5427 if (! rc) {
5428 *kernel = 0;
5429 *user = 0;
5430 return FALSE;
5431 }
5432 *kernel = (CHUNK_SIZE_T) (kernel64 / 10000);
5433 *user = (CHUNK_SIZE_T) (user64 / 10000);
5434 return TRUE;
5435 } else {
5436 __int64 creation64, exit64, kernel64, user64;
5437 int rc = GetThreadTimes (GetCurrentThread (),
5438 (FILETIME *) &creation64,
5439 (FILETIME *) &exit64,
5440 (FILETIME *) &kernel64,
5441 (FILETIME *) &user64);
5442 if (! rc) {
5443 *kernel = 0;
5444 *user = 0;
5445 return FALSE;
5446 }
5447 *kernel = (CHUNK_SIZE_T) (kernel64 / 10000);
5448 *user = (CHUNK_SIZE_T) (user64 / 10000);
5449 return TRUE;
5450 }
5451 }
5452
5453 #endif /* WIN32 */
5454
5455 /* ------------------------------------------------------------
5456 History:
5457 V2.7.2 Sat Aug 17 09:07:30 2002 Doug Lea (dl at gee)
5458 * Fix malloc_state bitmap array misdeclaration
5459
5460 V2.7.1 Thu Jul 25 10:58:03 2002 Doug Lea (dl at gee)
5461 * Allow tuning of FIRST_SORTED_BIN_SIZE
5462 * Use PTR_UINT as type for all ptr->int casts. Thanks to John Belmonte.
5463 * Better detection and support for non-contiguousness of MORECORE.
5464 Thanks to Andreas Mueller, Conal Walsh, and Wolfram Gloger
5465 * Bypass most of malloc if no frees. Thanks To Emery Berger.
5466 * Fix freeing of old top non-contiguous chunk im sysmalloc.
5467 * Raised default trim and map thresholds to 256K.
5468 * Fix mmap-related #defines. Thanks to Lubos Lunak.
5469 * Fix copy macros; added LACKS_FCNTL_H. Thanks to Neal Walfield.
5470 * Branch-free bin calculation
5471 * Default trim and mmap thresholds now 256K.
5472
5473 V2.7.0 Sun Mar 11 14:14:06 2001 Doug Lea (dl at gee)
5474 * Introduce independent_comalloc and independent_calloc.
5475 Thanks to Michael Pachos for motivation and help.
5476 * Make optional .h file available
5477 * Allow > 2GB requests on 32bit systems.
5478 * new WIN32 sbrk, mmap, munmap, lock code from <Walter@GeNeSys-e.de>.
5479 Thanks also to Andreas Mueller <a.mueller at paradatec.de>,
5480 and Anonymous.
5481 * Allow override of MALLOC_ALIGNMENT (Thanks to Ruud Waij for
5482 helping test this.)
5483 * memalign: check alignment arg
5484 * realloc: don't try to shift chunks backwards, since this
5485 leads to more fragmentation in some programs and doesn't
5486 seem to help in any others.
5487 * Collect all cases in malloc requiring system memory into sYSMALLOc
5488 * Use mmap as backup to sbrk
5489 * Place all internal state in malloc_state
5490 * Introduce fastbins (although similar to 2.5.1)
5491 * Many minor tunings and cosmetic improvements
5492 * Introduce USE_PUBLIC_MALLOC_WRAPPERS, USE_MALLOC_LOCK
5493 * Introduce MALLOC_FAILURE_ACTION, MORECORE_CONTIGUOUS
5494 Thanks to Tony E. Bennett <tbennett@nvidia.com> and others.
5495 * Include errno.h to support default failure action.
5496
5497 V2.6.6 Sun Dec 5 07:42:19 1999 Doug Lea (dl at gee)
5498 * return null for negative arguments
5499 * Added Several WIN32 cleanups from Martin C. Fong <mcfong at yahoo.com>
5500 * Add 'LACKS_SYS_PARAM_H' for those systems without 'sys/param.h'
5501 (e.g. WIN32 platforms)
5502 * Cleanup header file inclusion for WIN32 platforms
5503 * Cleanup code to avoid Microsoft Visual C++ compiler complaints
5504 * Add 'USE_DL_PREFIX' to quickly allow co-existence with existing
5505 memory allocation routines
5506 * Set 'malloc_getpagesize' for WIN32 platforms (needs more work)
5507 * Use 'assert' rather than 'ASSERT' in WIN32 code to conform to
5508 usage of 'assert' in non-WIN32 code
5509 * Improve WIN32 'sbrk()' emulation's 'findRegion()' routine to
5510 avoid infinite loop
5511 * Always call 'fREe()' rather than 'free()'
5512
5513 V2.6.5 Wed Jun 17 15:57:31 1998 Doug Lea (dl at gee)
5514 * Fixed ordering problem with boundary-stamping
5515
5516 V2.6.3 Sun May 19 08:17:58 1996 Doug Lea (dl at gee)
5517 * Added pvalloc, as recommended by H.J. Liu
5518 * Added 64bit pointer support mainly from Wolfram Gloger
5519 * Added anonymously donated WIN32 sbrk emulation
5520 * Malloc, calloc, getpagesize: add optimizations from Raymond Nijssen
5521 * malloc_extend_top: fix mask error that caused wastage after
5522 foreign sbrks
5523 * Add linux mremap support code from HJ Liu
5524
5525 V2.6.2 Tue Dec 5 06:52:55 1995 Doug Lea (dl at gee)
5526 * Integrated most documentation with the code.
5527 * Add support for mmap, with help from
5528 Wolfram Gloger (Gloger@lrz.uni-muenchen.de).
5529 * Use last_remainder in more cases.
5530 * Pack bins using idea from colin@nyx10.cs.du.edu
5531 * Use ordered bins instead of best-fit threshhold
5532 * Eliminate block-local decls to simplify tracing and debugging.
5533 * Support another case of realloc via move into top
5534 * Fix error occuring when initial sbrk_base not word-aligned.
5535 * Rely on page size for units instead of SBRK_UNIT to
5536 avoid surprises about sbrk alignment conventions.
5537 * Add mallinfo, mallopt. Thanks to Raymond Nijssen
5538 (raymond@es.ele.tue.nl) for the suggestion.
5539 * Add `pad' argument to malloc_trim and top_pad mallopt parameter.
5540 * More precautions for cases where other routines call sbrk,
5541 courtesy of Wolfram Gloger (Gloger@lrz.uni-muenchen.de).
5542 * Added macros etc., allowing use in linux libc from
5543 H.J. Lu (hjl@gnu.ai.mit.edu)
5544 * Inverted this history list
5545
5546 V2.6.1 Sat Dec 2 14:10:57 1995 Doug Lea (dl at gee)
5547 * Re-tuned and fixed to behave more nicely with V2.6.0 changes.
5548 * Removed all preallocation code since under current scheme
5549 the work required to undo bad preallocations exceeds
5550 the work saved in good cases for most test programs.
5551 * No longer use return list or unconsolidated bins since
5552 no scheme using them consistently outperforms those that don't
5553 given above changes.
5554 * Use best fit for very large chunks to prevent some worst-cases.
5555 * Added some support for debugging
5556
5557 V2.6.0 Sat Nov 4 07:05:23 1995 Doug Lea (dl at gee)
5558 * Removed footers when chunks are in use. Thanks to
5559 Paul Wilson (wilson@cs.texas.edu) for the suggestion.
5560
5561 V2.5.4 Wed Nov 1 07:54:51 1995 Doug Lea (dl at gee)
5562 * Added malloc_trim, with help from Wolfram Gloger
5563 (wmglo@Dent.MED.Uni-Muenchen.DE).
5564
5565 V2.5.3 Tue Apr 26 10:16:01 1994 Doug Lea (dl at g)
5566
5567 V2.5.2 Tue Apr 5 16:20:40 1994 Doug Lea (dl at g)
5568 * realloc: try to expand in both directions
5569 * malloc: swap order of clean-bin strategy;
5570 * realloc: only conditionally expand backwards
5571 * Try not to scavenge used bins
5572 * Use bin counts as a guide to preallocation
5573 * Occasionally bin return list chunks in first scan
5574 * Add a few optimizations from colin@nyx10.cs.du.edu
5575
5576 V2.5.1 Sat Aug 14 15:40:43 1993 Doug Lea (dl at g)
5577 * faster bin computation & slightly different binning
5578 * merged all consolidations to one part of malloc proper
5579 (eliminating old malloc_find_space & malloc_clean_bin)
5580 * Scan 2 returns chunks (not just 1)
5581 * Propagate failure in realloc if malloc returns 0
5582 * Add stuff to allow compilation on non-ANSI compilers
5583 from kpv@research.att.com
5584
5585 V2.5 Sat Aug 7 07:41:59 1993 Doug Lea (dl at g.oswego.edu)
5586 * removed potential for odd address access in prev_chunk
5587 * removed dependency on getpagesize.h
5588 * misc cosmetics and a bit more internal documentation
5589 * anticosmetics: mangled names in macros to evade debugger strangeness
5590 * tested on sparc, hp-700, dec-mips, rs6000
5591 with gcc & native cc (hp, dec only) allowing
5592 Detlefs & Zorn comparison study (in SIGPLAN Notices.)
5593
5594 Trial version Fri Aug 28 13:14:29 1992 Doug Lea (dl at g.oswego.edu)
5595 * Based loosely on libg++-1.2X malloc. (It retains some of the overall
5596 structure of old version, but most details differ.)
5597
5598 */