blob: e9bab09b8eac7c0ec4900a3685f15a807b448c31 [file] [log] [blame]
Kumar Gala81673e92008-05-13 19:01:54 -05001#include <common.h>
2
wdenk217c9da2002-10-25 20:35:49 +00003#if 0 /* Moved to malloc.h */
4/* ---------- To make a malloc.h, start cutting here ------------ */
5
6/*
7 A version of malloc/free/realloc written by Doug Lea and released to the
8 public domain. Send questions/comments/complaints/performance data
9 to dl@cs.oswego.edu
10
11* VERSION 2.6.6 Sun Mar 5 19:10:03 2000 Doug Lea (dl at gee)
12
13 Note: There may be an updated version of this malloc obtainable at
wdenk8bde7f72003-06-27 21:31:46 +000014 ftp://g.oswego.edu/pub/misc/malloc.c
15 Check before installing!
wdenk217c9da2002-10-25 20:35:49 +000016
17* Why use this malloc?
18
19 This is not the fastest, most space-conserving, most portable, or
20 most tunable malloc ever written. However it is among the fastest
21 while also being among the most space-conserving, portable and tunable.
22 Consistent balance across these factors results in a good general-purpose
23 allocator. For a high-level description, see
24 http://g.oswego.edu/dl/html/malloc.html
25
26* Synopsis of public routines
27
28 (Much fuller descriptions are contained in the program documentation below.)
29
30 malloc(size_t n);
31 Return a pointer to a newly allocated chunk of at least n bytes, or null
32 if no space is available.
33 free(Void_t* p);
34 Release the chunk of memory pointed to by p, or no effect if p is null.
35 realloc(Void_t* p, size_t n);
36 Return a pointer to a chunk of size n that contains the same data
37 as does chunk p up to the minimum of (n, p's size) bytes, or null
38 if no space is available. The returned pointer may or may not be
39 the same as p. If p is null, equivalent to malloc. Unless the
40 #define REALLOC_ZERO_BYTES_FREES below is set, realloc with a
41 size argument of zero (re)allocates a minimum-sized chunk.
42 memalign(size_t alignment, size_t n);
43 Return a pointer to a newly allocated chunk of n bytes, aligned
44 in accord with the alignment argument, which must be a power of
45 two.
46 valloc(size_t n);
47 Equivalent to memalign(pagesize, n), where pagesize is the page
48 size of the system (or as near to this as can be figured out from
49 all the includes/defines below.)
50 pvalloc(size_t n);
51 Equivalent to valloc(minimum-page-that-holds(n)), that is,
52 round up n to nearest pagesize.
53 calloc(size_t unit, size_t quantity);
54 Returns a pointer to quantity * unit bytes, with all locations
55 set to zero.
56 cfree(Void_t* p);
57 Equivalent to free(p).
58 malloc_trim(size_t pad);
59 Release all but pad bytes of freed top-most memory back
60 to the system. Return 1 if successful, else 0.
61 malloc_usable_size(Void_t* p);
62 Report the number usable allocated bytes associated with allocated
63 chunk p. This may or may not report more bytes than were requested,
64 due to alignment and minimum size constraints.
65 malloc_stats();
66 Prints brief summary statistics.
67 mallinfo()
68 Returns (by copy) a struct containing various summary statistics.
69 mallopt(int parameter_number, int parameter_value)
70 Changes one of the tunable parameters described below. Returns
71 1 if successful in changing the parameter, else 0.
72
73* Vital statistics:
74
75 Alignment: 8-byte
76 8 byte alignment is currently hardwired into the design. This
77 seems to suffice for all current machines and C compilers.
78
79 Assumed pointer representation: 4 or 8 bytes
80 Code for 8-byte pointers is untested by me but has worked
81 reliably by Wolfram Gloger, who contributed most of the
82 changes supporting this.
83
84 Assumed size_t representation: 4 or 8 bytes
85 Note that size_t is allowed to be 4 bytes even if pointers are 8.
86
87 Minimum overhead per allocated chunk: 4 or 8 bytes
88 Each malloced chunk has a hidden overhead of 4 bytes holding size
89 and status information.
90
91 Minimum allocated size: 4-byte ptrs: 16 bytes (including 4 overhead)
wdenk8bde7f72003-06-27 21:31:46 +000092 8-byte ptrs: 24/32 bytes (including, 4/8 overhead)
wdenk217c9da2002-10-25 20:35:49 +000093
94 When a chunk is freed, 12 (for 4byte ptrs) or 20 (for 8 byte
95 ptrs but 4 byte size) or 24 (for 8/8) additional bytes are
96 needed; 4 (8) for a trailing size field
97 and 8 (16) bytes for free list pointers. Thus, the minimum
98 allocatable size is 16/24/32 bytes.
99
100 Even a request for zero bytes (i.e., malloc(0)) returns a
101 pointer to something of the minimum allocatable size.
102
103 Maximum allocated size: 4-byte size_t: 2^31 - 8 bytes
wdenk8bde7f72003-06-27 21:31:46 +0000104 8-byte size_t: 2^63 - 16 bytes
wdenk217c9da2002-10-25 20:35:49 +0000105
106 It is assumed that (possibly signed) size_t bit values suffice to
107 represent chunk sizes. `Possibly signed' is due to the fact
108 that `size_t' may be defined on a system as either a signed or
109 an unsigned type. To be conservative, values that would appear
110 as negative numbers are avoided.
111 Requests for sizes with a negative sign bit when the request
112 size is treaded as a long will return null.
113
114 Maximum overhead wastage per allocated chunk: normally 15 bytes
115
116 Alignnment demands, plus the minimum allocatable size restriction
117 make the normal worst-case wastage 15 bytes (i.e., up to 15
118 more bytes will be allocated than were requested in malloc), with
119 two exceptions:
wdenk8bde7f72003-06-27 21:31:46 +0000120 1. Because requests for zero bytes allocate non-zero space,
121 the worst case wastage for a request of zero bytes is 24 bytes.
122 2. For requests >= mmap_threshold that are serviced via
123 mmap(), the worst case wastage is 8 bytes plus the remainder
124 from a system page (the minimal mmap unit); typically 4096 bytes.
wdenk217c9da2002-10-25 20:35:49 +0000125
126* Limitations
127
128 Here are some features that are NOT currently supported
129
130 * No user-definable hooks for callbacks and the like.
131 * No automated mechanism for fully checking that all accesses
132 to malloced memory stay within their bounds.
133 * No support for compaction.
134
135* Synopsis of compile-time options:
136
137 People have reported using previous versions of this malloc on all
138 versions of Unix, sometimes by tweaking some of the defines
139 below. It has been tested most extensively on Solaris and
140 Linux. It is also reported to work on WIN32 platforms.
141 People have also reported adapting this malloc for use in
142 stand-alone embedded systems.
143
144 The implementation is in straight, hand-tuned ANSI C. Among other
145 consequences, it uses a lot of macros. Because of this, to be at
146 all usable, this code should be compiled using an optimizing compiler
147 (for example gcc -O2) that can simplify expressions and control
148 paths.
149
150 __STD_C (default: derived from C compiler defines)
151 Nonzero if using ANSI-standard C compiler, a C++ compiler, or
152 a C compiler sufficiently close to ANSI to get away with it.
153 DEBUG (default: NOT defined)
154 Define to enable debugging. Adds fairly extensive assertion-based
155 checking to help track down memory errors, but noticeably slows down
156 execution.
157 REALLOC_ZERO_BYTES_FREES (default: NOT defined)
158 Define this if you think that realloc(p, 0) should be equivalent
159 to free(p). Otherwise, since malloc returns a unique pointer for
160 malloc(0), so does realloc(p, 0).
161 HAVE_MEMCPY (default: defined)
162 Define if you are not otherwise using ANSI STD C, but still
163 have memcpy and memset in your C library and want to use them.
164 Otherwise, simple internal versions are supplied.
165 USE_MEMCPY (default: 1 if HAVE_MEMCPY is defined, 0 otherwise)
166 Define as 1 if you want the C library versions of memset and
167 memcpy called in realloc and calloc (otherwise macro versions are used).
168 At least on some platforms, the simple macro versions usually
169 outperform libc versions.
170 HAVE_MMAP (default: defined as 1)
171 Define to non-zero to optionally make malloc() use mmap() to
172 allocate very large blocks.
173 HAVE_MREMAP (default: defined as 0 unless Linux libc set)
174 Define to non-zero to optionally make realloc() use mremap() to
175 reallocate very large blocks.
176 malloc_getpagesize (default: derived from system #includes)
177 Either a constant or routine call returning the system page size.
178 HAVE_USR_INCLUDE_MALLOC_H (default: NOT defined)
179 Optionally define if you are on a system with a /usr/include/malloc.h
180 that declares struct mallinfo. It is not at all necessary to
181 define this even if you do, but will ensure consistency.
182 INTERNAL_SIZE_T (default: size_t)
183 Define to a 32-bit type (probably `unsigned int') if you are on a
184 64-bit machine, yet do not want or need to allow malloc requests of
185 greater than 2^31 to be handled. This saves space, especially for
186 very small chunks.
187 INTERNAL_LINUX_C_LIB (default: NOT defined)
188 Defined only when compiled as part of Linux libc.
189 Also note that there is some odd internal name-mangling via defines
190 (for example, internally, `malloc' is named `mALLOc') needed
191 when compiling in this case. These look funny but don't otherwise
192 affect anything.
193 WIN32 (default: undefined)
194 Define this on MS win (95, nt) platforms to compile in sbrk emulation.
195 LACKS_UNISTD_H (default: undefined if not WIN32)
196 Define this if your system does not have a <unistd.h>.
197 LACKS_SYS_PARAM_H (default: undefined if not WIN32)
198 Define this if your system does not have a <sys/param.h>.
199 MORECORE (default: sbrk)
200 The name of the routine to call to obtain more memory from the system.
201 MORECORE_FAILURE (default: -1)
202 The value returned upon failure of MORECORE.
203 MORECORE_CLEARS (default 1)
204 True (1) if the routine mapped to MORECORE zeroes out memory (which
205 holds for sbrk).
206 DEFAULT_TRIM_THRESHOLD
207 DEFAULT_TOP_PAD
208 DEFAULT_MMAP_THRESHOLD
209 DEFAULT_MMAP_MAX
210 Default values of tunable parameters (described in detail below)
211 controlling interaction with host system routines (sbrk, mmap, etc).
212 These values may also be changed dynamically via mallopt(). The
213 preset defaults are those that give best performance for typical
214 programs/systems.
215 USE_DL_PREFIX (default: undefined)
216 Prefix all public routines with the string 'dl'. Useful to
217 quickly avoid procedure declaration conflicts and linker symbol
218 conflicts with existing memory allocation routines.
219
220
221*/
222
223
224
wdenk217c9da2002-10-25 20:35:49 +0000225/* Preliminaries */
226
227#ifndef __STD_C
228#ifdef __STDC__
229#define __STD_C 1
230#else
231#if __cplusplus
232#define __STD_C 1
233#else
234#define __STD_C 0
235#endif /*__cplusplus*/
236#endif /*__STDC__*/
237#endif /*__STD_C*/
238
239#ifndef Void_t
240#if (__STD_C || defined(WIN32))
241#define Void_t void
242#else
243#define Void_t char
244#endif
245#endif /*Void_t*/
246
247#if __STD_C
248#include <stddef.h> /* for size_t */
249#else
250#include <sys/types.h>
251#endif
252
253#ifdef __cplusplus
254extern "C" {
255#endif
256
257#include <stdio.h> /* needed for malloc_stats */
258
259
260/*
261 Compile-time options
262*/
263
264
265/*
266 Debugging:
267
268 Because freed chunks may be overwritten with link fields, this
269 malloc will often die when freed memory is overwritten by user
270 programs. This can be very effective (albeit in an annoying way)
271 in helping track down dangling pointers.
272
273 If you compile with -DDEBUG, a number of assertion checks are
274 enabled that will catch more memory errors. You probably won't be
275 able to make much sense of the actual assertion errors, but they
276 should help you locate incorrectly overwritten memory. The
277 checking is fairly extensive, and will slow down execution
278 noticeably. Calling malloc_stats or mallinfo with DEBUG set will
279 attempt to check every non-mmapped allocated and free chunk in the
280 course of computing the summmaries. (By nature, mmapped regions
281 cannot be checked very much automatically.)
282
283 Setting DEBUG may also be helpful if you are trying to modify
284 this code. The assertions in the check routines spell out in more
285 detail the assumptions and invariants underlying the algorithms.
286
287*/
288
289#ifdef DEBUG
290#include <assert.h>
291#else
292#define assert(x) ((void)0)
293#endif
294
295
296/*
297 INTERNAL_SIZE_T is the word-size used for internal bookkeeping
298 of chunk sizes. On a 64-bit machine, you can reduce malloc
299 overhead by defining INTERNAL_SIZE_T to be a 32 bit `unsigned int'
300 at the expense of not being able to handle requests greater than
301 2^31. This limitation is hardly ever a concern; you are encouraged
302 to set this. However, the default version is the same as size_t.
303*/
304
305#ifndef INTERNAL_SIZE_T
306#define INTERNAL_SIZE_T size_t
307#endif
308
309/*
310 REALLOC_ZERO_BYTES_FREES should be set if a call to
311 realloc with zero bytes should be the same as a call to free.
312 Some people think it should. Otherwise, since this malloc
313 returns a unique pointer for malloc(0), so does realloc(p, 0).
314*/
315
316
317/* #define REALLOC_ZERO_BYTES_FREES */
318
319
320/*
321 WIN32 causes an emulation of sbrk to be compiled in
322 mmap-based options are not currently supported in WIN32.
323*/
324
325/* #define WIN32 */
326#ifdef WIN32
327#define MORECORE wsbrk
328#define HAVE_MMAP 0
329
330#define LACKS_UNISTD_H
331#define LACKS_SYS_PARAM_H
332
333/*
334 Include 'windows.h' to get the necessary declarations for the
335 Microsoft Visual C++ data structures and routines used in the 'sbrk'
336 emulation.
337
338 Define WIN32_LEAN_AND_MEAN so that only the essential Microsoft
339 Visual C++ header files are included.
340*/
341#define WIN32_LEAN_AND_MEAN
342#include <windows.h>
343#endif
344
345
346/*
347 HAVE_MEMCPY should be defined if you are not otherwise using
348 ANSI STD C, but still have memcpy and memset in your C library
349 and want to use them in calloc and realloc. Otherwise simple
350 macro versions are defined here.
351
352 USE_MEMCPY should be defined as 1 if you actually want to
353 have memset and memcpy called. People report that the macro
354 versions are often enough faster than libc versions on many
355 systems that it is better to use them.
356
357*/
358
359#define HAVE_MEMCPY
360
361#ifndef USE_MEMCPY
362#ifdef HAVE_MEMCPY
363#define USE_MEMCPY 1
364#else
365#define USE_MEMCPY 0
366#endif
367#endif
368
369#if (__STD_C || defined(HAVE_MEMCPY))
370
371#if __STD_C
372void* memset(void*, int, size_t);
373void* memcpy(void*, const void*, size_t);
374#else
375#ifdef WIN32
wdenk8bde7f72003-06-27 21:31:46 +0000376/* On Win32 platforms, 'memset()' and 'memcpy()' are already declared in */
377/* 'windows.h' */
wdenk217c9da2002-10-25 20:35:49 +0000378#else
379Void_t* memset();
380Void_t* memcpy();
381#endif
382#endif
383#endif
384
385#if USE_MEMCPY
386
387/* The following macros are only invoked with (2n+1)-multiples of
388 INTERNAL_SIZE_T units, with a positive integer n. This is exploited
389 for fast inline execution when n is small. */
390
391#define MALLOC_ZERO(charp, nbytes) \
392do { \
393 INTERNAL_SIZE_T mzsz = (nbytes); \
394 if(mzsz <= 9*sizeof(mzsz)) { \
395 INTERNAL_SIZE_T* mz = (INTERNAL_SIZE_T*) (charp); \
396 if(mzsz >= 5*sizeof(mzsz)) { *mz++ = 0; \
wdenk8bde7f72003-06-27 21:31:46 +0000397 *mz++ = 0; \
wdenk217c9da2002-10-25 20:35:49 +0000398 if(mzsz >= 7*sizeof(mzsz)) { *mz++ = 0; \
wdenk8bde7f72003-06-27 21:31:46 +0000399 *mz++ = 0; \
400 if(mzsz >= 9*sizeof(mzsz)) { *mz++ = 0; \
401 *mz++ = 0; }}} \
402 *mz++ = 0; \
403 *mz++ = 0; \
404 *mz = 0; \
wdenk217c9da2002-10-25 20:35:49 +0000405 } else memset((charp), 0, mzsz); \
406} while(0)
407
408#define MALLOC_COPY(dest,src,nbytes) \
409do { \
410 INTERNAL_SIZE_T mcsz = (nbytes); \
411 if(mcsz <= 9*sizeof(mcsz)) { \
412 INTERNAL_SIZE_T* mcsrc = (INTERNAL_SIZE_T*) (src); \
413 INTERNAL_SIZE_T* mcdst = (INTERNAL_SIZE_T*) (dest); \
414 if(mcsz >= 5*sizeof(mcsz)) { *mcdst++ = *mcsrc++; \
wdenk8bde7f72003-06-27 21:31:46 +0000415 *mcdst++ = *mcsrc++; \
wdenk217c9da2002-10-25 20:35:49 +0000416 if(mcsz >= 7*sizeof(mcsz)) { *mcdst++ = *mcsrc++; \
wdenk8bde7f72003-06-27 21:31:46 +0000417 *mcdst++ = *mcsrc++; \
418 if(mcsz >= 9*sizeof(mcsz)) { *mcdst++ = *mcsrc++; \
419 *mcdst++ = *mcsrc++; }}} \
420 *mcdst++ = *mcsrc++; \
421 *mcdst++ = *mcsrc++; \
422 *mcdst = *mcsrc ; \
wdenk217c9da2002-10-25 20:35:49 +0000423 } else memcpy(dest, src, mcsz); \
424} while(0)
425
426#else /* !USE_MEMCPY */
427
428/* Use Duff's device for good zeroing/copying performance. */
429
430#define MALLOC_ZERO(charp, nbytes) \
431do { \
432 INTERNAL_SIZE_T* mzp = (INTERNAL_SIZE_T*)(charp); \
433 long mctmp = (nbytes)/sizeof(INTERNAL_SIZE_T), mcn; \
434 if (mctmp < 8) mcn = 0; else { mcn = (mctmp-1)/8; mctmp %= 8; } \
435 switch (mctmp) { \
436 case 0: for(;;) { *mzp++ = 0; \
437 case 7: *mzp++ = 0; \
438 case 6: *mzp++ = 0; \
439 case 5: *mzp++ = 0; \
440 case 4: *mzp++ = 0; \
441 case 3: *mzp++ = 0; \
442 case 2: *mzp++ = 0; \
443 case 1: *mzp++ = 0; if(mcn <= 0) break; mcn--; } \
444 } \
445} while(0)
446
447#define MALLOC_COPY(dest,src,nbytes) \
448do { \
449 INTERNAL_SIZE_T* mcsrc = (INTERNAL_SIZE_T*) src; \
450 INTERNAL_SIZE_T* mcdst = (INTERNAL_SIZE_T*) dest; \
451 long mctmp = (nbytes)/sizeof(INTERNAL_SIZE_T), mcn; \
452 if (mctmp < 8) mcn = 0; else { mcn = (mctmp-1)/8; mctmp %= 8; } \
453 switch (mctmp) { \
454 case 0: for(;;) { *mcdst++ = *mcsrc++; \
455 case 7: *mcdst++ = *mcsrc++; \
456 case 6: *mcdst++ = *mcsrc++; \
457 case 5: *mcdst++ = *mcsrc++; \
458 case 4: *mcdst++ = *mcsrc++; \
459 case 3: *mcdst++ = *mcsrc++; \
460 case 2: *mcdst++ = *mcsrc++; \
461 case 1: *mcdst++ = *mcsrc++; if(mcn <= 0) break; mcn--; } \
462 } \
463} while(0)
464
465#endif
466
467
468/*
469 Define HAVE_MMAP to optionally make malloc() use mmap() to
470 allocate very large blocks. These will be returned to the
471 operating system immediately after a free().
472*/
473
474#ifndef HAVE_MMAP
475#define HAVE_MMAP 1
476#endif
477
478/*
479 Define HAVE_MREMAP to make realloc() use mremap() to re-allocate
480 large blocks. This is currently only possible on Linux with
481 kernel versions newer than 1.3.77.
482*/
483
484#ifndef HAVE_MREMAP
485#ifdef INTERNAL_LINUX_C_LIB
486#define HAVE_MREMAP 1
487#else
488#define HAVE_MREMAP 0
489#endif
490#endif
491
492#if HAVE_MMAP
493
494#include <unistd.h>
495#include <fcntl.h>
496#include <sys/mman.h>
497
498#if !defined(MAP_ANONYMOUS) && defined(MAP_ANON)
499#define MAP_ANONYMOUS MAP_ANON
500#endif
501
502#endif /* HAVE_MMAP */
503
504/*
505 Access to system page size. To the extent possible, this malloc
506 manages memory from the system in page-size units.
507
508 The following mechanics for getpagesize were adapted from
509 bsd/gnu getpagesize.h
510*/
511
512#ifndef LACKS_UNISTD_H
513# include <unistd.h>
514#endif
515
516#ifndef malloc_getpagesize
517# ifdef _SC_PAGESIZE /* some SVR4 systems omit an underscore */
518# ifndef _SC_PAGE_SIZE
519# define _SC_PAGE_SIZE _SC_PAGESIZE
520# endif
521# endif
522# ifdef _SC_PAGE_SIZE
523# define malloc_getpagesize sysconf(_SC_PAGE_SIZE)
524# else
525# if defined(BSD) || defined(DGUX) || defined(HAVE_GETPAGESIZE)
526 extern size_t getpagesize();
527# define malloc_getpagesize getpagesize()
528# else
529# ifdef WIN32
530# define malloc_getpagesize (4096) /* TBD: Use 'GetSystemInfo' instead */
531# else
532# ifndef LACKS_SYS_PARAM_H
533# include <sys/param.h>
534# endif
535# ifdef EXEC_PAGESIZE
536# define malloc_getpagesize EXEC_PAGESIZE
537# else
538# ifdef NBPG
539# ifndef CLSIZE
540# define malloc_getpagesize NBPG
541# else
542# define malloc_getpagesize (NBPG * CLSIZE)
543# endif
544# else
545# ifdef NBPC
546# define malloc_getpagesize NBPC
547# else
548# ifdef PAGESIZE
549# define malloc_getpagesize PAGESIZE
550# else
551# define malloc_getpagesize (4096) /* just guess */
552# endif
553# endif
554# endif
555# endif
556# endif
557# endif
558# endif
559#endif
560
561
wdenk217c9da2002-10-25 20:35:49 +0000562/*
563
564 This version of malloc supports the standard SVID/XPG mallinfo
565 routine that returns a struct containing the same kind of
566 information you can get from malloc_stats. It should work on
567 any SVID/XPG compliant system that has a /usr/include/malloc.h
568 defining struct mallinfo. (If you'd like to install such a thing
569 yourself, cut out the preliminary declarations as described above
570 and below and save them in a malloc.h file. But there's no
571 compelling reason to bother to do this.)
572
573 The main declaration needed is the mallinfo struct that is returned
574 (by-copy) by mallinfo(). The SVID/XPG malloinfo struct contains a
575 bunch of fields, most of which are not even meaningful in this
576 version of malloc. Some of these fields are are instead filled by
577 mallinfo() with other numbers that might possibly be of interest.
578
579 HAVE_USR_INCLUDE_MALLOC_H should be set if you have a
580 /usr/include/malloc.h file that includes a declaration of struct
581 mallinfo. If so, it is included; else an SVID2/XPG2 compliant
582 version is declared below. These must be precisely the same for
583 mallinfo() to work.
584
585*/
586
587/* #define HAVE_USR_INCLUDE_MALLOC_H */
588
589#if HAVE_USR_INCLUDE_MALLOC_H
590#include "/usr/include/malloc.h"
591#else
592
593/* SVID2/XPG mallinfo structure */
594
595struct mallinfo {
596 int arena; /* total space allocated from system */
597 int ordblks; /* number of non-inuse chunks */
598 int smblks; /* unused -- always zero */
599 int hblks; /* number of mmapped regions */
600 int hblkhd; /* total space in mmapped regions */
601 int usmblks; /* unused -- always zero */
602 int fsmblks; /* unused -- always zero */
603 int uordblks; /* total allocated space */
604 int fordblks; /* total non-inuse space */
605 int keepcost; /* top-most, releasable (via malloc_trim) space */
606};
607
608/* SVID2/XPG mallopt options */
609
610#define M_MXFAST 1 /* UNUSED in this malloc */
611#define M_NLBLKS 2 /* UNUSED in this malloc */
612#define M_GRAIN 3 /* UNUSED in this malloc */
613#define M_KEEP 4 /* UNUSED in this malloc */
614
615#endif
616
617/* mallopt options that actually do something */
618
619#define M_TRIM_THRESHOLD -1
620#define M_TOP_PAD -2
621#define M_MMAP_THRESHOLD -3
622#define M_MMAP_MAX -4
623
624
wdenk217c9da2002-10-25 20:35:49 +0000625#ifndef DEFAULT_TRIM_THRESHOLD
626#define DEFAULT_TRIM_THRESHOLD (128 * 1024)
627#endif
628
629/*
630 M_TRIM_THRESHOLD is the maximum amount of unused top-most memory
631 to keep before releasing via malloc_trim in free().
632
633 Automatic trimming is mainly useful in long-lived programs.
634 Because trimming via sbrk can be slow on some systems, and can
635 sometimes be wasteful (in cases where programs immediately
636 afterward allocate more large chunks) the value should be high
637 enough so that your overall system performance would improve by
638 releasing.
639
640 The trim threshold and the mmap control parameters (see below)
641 can be traded off with one another. Trimming and mmapping are
642 two different ways of releasing unused memory back to the
643 system. Between these two, it is often possible to keep
644 system-level demands of a long-lived program down to a bare
645 minimum. For example, in one test suite of sessions measuring
646 the XF86 X server on Linux, using a trim threshold of 128K and a
647 mmap threshold of 192K led to near-minimal long term resource
648 consumption.
649
650 If you are using this malloc in a long-lived program, it should
651 pay to experiment with these values. As a rough guide, you
652 might set to a value close to the average size of a process
653 (program) running on your system. Releasing this much memory
654 would allow such a process to run in memory. Generally, it's
655 worth it to tune for trimming rather tham memory mapping when a
656 program undergoes phases where several large chunks are
657 allocated and released in ways that can reuse each other's
658 storage, perhaps mixed with phases where there are no such
659 chunks at all. And in well-behaved long-lived programs,
660 controlling release of large blocks via trimming versus mapping
661 is usually faster.
662
663 However, in most programs, these parameters serve mainly as
664 protection against the system-level effects of carrying around
665 massive amounts of unneeded memory. Since frequent calls to
666 sbrk, mmap, and munmap otherwise degrade performance, the default
667 parameters are set to relatively high values that serve only as
668 safeguards.
669
670 The default trim value is high enough to cause trimming only in
671 fairly extreme (by current memory consumption standards) cases.
672 It must be greater than page size to have any useful effect. To
673 disable trimming completely, you can set to (unsigned long)(-1);
674
675
676*/
677
678
679#ifndef DEFAULT_TOP_PAD
680#define DEFAULT_TOP_PAD (0)
681#endif
682
683/*
684 M_TOP_PAD is the amount of extra `padding' space to allocate or
685 retain whenever sbrk is called. It is used in two ways internally:
686
687 * When sbrk is called to extend the top of the arena to satisfy
wdenk8bde7f72003-06-27 21:31:46 +0000688 a new malloc request, this much padding is added to the sbrk
689 request.
wdenk217c9da2002-10-25 20:35:49 +0000690
691 * When malloc_trim is called automatically from free(),
wdenk8bde7f72003-06-27 21:31:46 +0000692 it is used as the `pad' argument.
wdenk217c9da2002-10-25 20:35:49 +0000693
694 In both cases, the actual amount of padding is rounded
695 so that the end of the arena is always a system page boundary.
696
697 The main reason for using padding is to avoid calling sbrk so
698 often. Having even a small pad greatly reduces the likelihood
699 that nearly every malloc request during program start-up (or
700 after trimming) will invoke sbrk, which needlessly wastes
701 time.
702
703 Automatic rounding-up to page-size units is normally sufficient
704 to avoid measurable overhead, so the default is 0. However, in
705 systems where sbrk is relatively slow, it can pay to increase
706 this value, at the expense of carrying around more memory than
707 the program needs.
708
709*/
710
711
712#ifndef DEFAULT_MMAP_THRESHOLD
713#define DEFAULT_MMAP_THRESHOLD (128 * 1024)
714#endif
715
716/*
717
718 M_MMAP_THRESHOLD is the request size threshold for using mmap()
719 to service a request. Requests of at least this size that cannot
720 be allocated using already-existing space will be serviced via mmap.
721 (If enough normal freed space already exists it is used instead.)
722
723 Using mmap segregates relatively large chunks of memory so that
724 they can be individually obtained and released from the host
725 system. A request serviced through mmap is never reused by any
726 other request (at least not directly; the system may just so
727 happen to remap successive requests to the same locations).
728
729 Segregating space in this way has the benefit that mmapped space
730 can ALWAYS be individually released back to the system, which
731 helps keep the system level memory demands of a long-lived
732 program low. Mapped memory can never become `locked' between
733 other chunks, as can happen with normally allocated chunks, which
734 menas that even trimming via malloc_trim would not release them.
735
736 However, it has the disadvantages that:
737
wdenk8bde7f72003-06-27 21:31:46 +0000738 1. The space cannot be reclaimed, consolidated, and then
739 used to service later requests, as happens with normal chunks.
740 2. It can lead to more wastage because of mmap page alignment
741 requirements
742 3. It causes malloc performance to be more dependent on host
743 system memory management support routines which may vary in
744 implementation quality and may impose arbitrary
745 limitations. Generally, servicing a request via normal
746 malloc steps is faster than going through a system's mmap.
wdenk217c9da2002-10-25 20:35:49 +0000747
748 All together, these considerations should lead you to use mmap
749 only for relatively large requests.
750
751
752*/
753
754
wdenk217c9da2002-10-25 20:35:49 +0000755#ifndef DEFAULT_MMAP_MAX
756#if HAVE_MMAP
757#define DEFAULT_MMAP_MAX (64)
758#else
759#define DEFAULT_MMAP_MAX (0)
760#endif
761#endif
762
763/*
764 M_MMAP_MAX is the maximum number of requests to simultaneously
765 service using mmap. This parameter exists because:
766
wdenk8bde7f72003-06-27 21:31:46 +0000767 1. Some systems have a limited number of internal tables for
768 use by mmap.
769 2. In most systems, overreliance on mmap can degrade overall
770 performance.
771 3. If a program allocates many large regions, it is probably
772 better off using normal sbrk-based allocation routines that
773 can reclaim and reallocate normal heap memory. Using a
774 small value allows transition into this mode after the
775 first few allocations.
wdenk217c9da2002-10-25 20:35:49 +0000776
777 Setting to 0 disables all use of mmap. If HAVE_MMAP is not set,
778 the default value is 0, and attempts to set it to non-zero values
779 in mallopt will fail.
780*/
781
782
wdenk217c9da2002-10-25 20:35:49 +0000783/*
784 USE_DL_PREFIX will prefix all public routines with the string 'dl'.
785 Useful to quickly avoid procedure declaration conflicts and linker
786 symbol conflicts with existing memory allocation routines.
787
788*/
789
790/* #define USE_DL_PREFIX */
791
792
wdenk217c9da2002-10-25 20:35:49 +0000793/*
794
795 Special defines for linux libc
796
797 Except when compiled using these special defines for Linux libc
798 using weak aliases, this malloc is NOT designed to work in
799 multithreaded applications. No semaphores or other concurrency
800 control are provided to ensure that multiple malloc or free calls
801 don't run at the same time, which could be disasterous. A single
802 semaphore could be used across malloc, realloc, and free (which is
803 essentially the effect of the linux weak alias approach). It would
804 be hard to obtain finer granularity.
805
806*/
807
808
809#ifdef INTERNAL_LINUX_C_LIB
810
811#if __STD_C
812
813Void_t * __default_morecore_init (ptrdiff_t);
814Void_t *(*__morecore)(ptrdiff_t) = __default_morecore_init;
815
816#else
817
818Void_t * __default_morecore_init ();
819Void_t *(*__morecore)() = __default_morecore_init;
820
821#endif
822
823#define MORECORE (*__morecore)
824#define MORECORE_FAILURE 0
825#define MORECORE_CLEARS 1
826
827#else /* INTERNAL_LINUX_C_LIB */
828
829#if __STD_C
830extern Void_t* sbrk(ptrdiff_t);
831#else
832extern Void_t* sbrk();
833#endif
834
835#ifndef MORECORE
836#define MORECORE sbrk
837#endif
838
839#ifndef MORECORE_FAILURE
840#define MORECORE_FAILURE -1
841#endif
842
843#ifndef MORECORE_CLEARS
844#define MORECORE_CLEARS 1
845#endif
846
847#endif /* INTERNAL_LINUX_C_LIB */
848
849#if defined(INTERNAL_LINUX_C_LIB) && defined(__ELF__)
850
851#define cALLOc __libc_calloc
852#define fREe __libc_free
853#define mALLOc __libc_malloc
854#define mEMALIGn __libc_memalign
855#define rEALLOc __libc_realloc
856#define vALLOc __libc_valloc
857#define pvALLOc __libc_pvalloc
858#define mALLINFo __libc_mallinfo
859#define mALLOPt __libc_mallopt
860
861#pragma weak calloc = __libc_calloc
862#pragma weak free = __libc_free
863#pragma weak cfree = __libc_free
864#pragma weak malloc = __libc_malloc
865#pragma weak memalign = __libc_memalign
866#pragma weak realloc = __libc_realloc
867#pragma weak valloc = __libc_valloc
868#pragma weak pvalloc = __libc_pvalloc
869#pragma weak mallinfo = __libc_mallinfo
870#pragma weak mallopt = __libc_mallopt
871
872#else
873
874#ifdef USE_DL_PREFIX
875#define cALLOc dlcalloc
876#define fREe dlfree
877#define mALLOc dlmalloc
878#define mEMALIGn dlmemalign
879#define rEALLOc dlrealloc
880#define vALLOc dlvalloc
881#define pvALLOc dlpvalloc
882#define mALLINFo dlmallinfo
883#define mALLOPt dlmallopt
884#else /* USE_DL_PREFIX */
885#define cALLOc calloc
886#define fREe free
887#define mALLOc malloc
888#define mEMALIGn memalign
889#define rEALLOc realloc
890#define vALLOc valloc
891#define pvALLOc pvalloc
892#define mALLINFo mallinfo
893#define mALLOPt mallopt
894#endif /* USE_DL_PREFIX */
895
896#endif
897
898/* Public routines */
899
900#if __STD_C
901
902Void_t* mALLOc(size_t);
903void fREe(Void_t*);
904Void_t* rEALLOc(Void_t*, size_t);
905Void_t* mEMALIGn(size_t, size_t);
906Void_t* vALLOc(size_t);
907Void_t* pvALLOc(size_t);
908Void_t* cALLOc(size_t, size_t);
909void cfree(Void_t*);
910int malloc_trim(size_t);
911size_t malloc_usable_size(Void_t*);
912void malloc_stats();
913int mALLOPt(int, int);
914struct mallinfo mALLINFo(void);
915#else
916Void_t* mALLOc();
917void fREe();
918Void_t* rEALLOc();
919Void_t* mEMALIGn();
920Void_t* vALLOc();
921Void_t* pvALLOc();
922Void_t* cALLOc();
923void cfree();
924int malloc_trim();
925size_t malloc_usable_size();
926void malloc_stats();
927int mALLOPt();
928struct mallinfo mALLINFo();
929#endif
930
931
932#ifdef __cplusplus
933}; /* end of extern "C" */
934#endif
935
936/* ---------- To make a malloc.h, end cutting here ------------ */
Wolfgang Denkea882ba2010-06-20 23:33:59 +0200937#endif /* 0 */ /* Moved to malloc.h */
wdenk217c9da2002-10-25 20:35:49 +0000938
939#include <malloc.h>
Wolfgang Denkea882ba2010-06-20 23:33:59 +0200940#ifdef DEBUG
wdenk217c9da2002-10-25 20:35:49 +0000941#if __STD_C
942static void malloc_update_mallinfo (void);
943void malloc_stats (void);
944#else
945static void malloc_update_mallinfo ();
946void malloc_stats();
947#endif
Wolfgang Denkea882ba2010-06-20 23:33:59 +0200948#endif /* DEBUG */
wdenk217c9da2002-10-25 20:35:49 +0000949
Wolfgang Denkd87080b2006-03-31 18:32:53 +0200950DECLARE_GLOBAL_DATA_PTR;
951
wdenk217c9da2002-10-25 20:35:49 +0000952/*
953 Emulation of sbrk for WIN32
954 All code within the ifdef WIN32 is untested by me.
955
956 Thanks to Martin Fong and others for supplying this.
957*/
958
959
960#ifdef WIN32
961
962#define AlignPage(add) (((add) + (malloc_getpagesize-1)) & \
963~(malloc_getpagesize-1))
964#define AlignPage64K(add) (((add) + (0x10000 - 1)) & ~(0x10000 - 1))
965
966/* resrve 64MB to insure large contiguous space */
967#define RESERVED_SIZE (1024*1024*64)
968#define NEXT_SIZE (2048*1024)
969#define TOP_MEMORY ((unsigned long)2*1024*1024*1024)
970
971struct GmListElement;
972typedef struct GmListElement GmListElement;
973
974struct GmListElement
975{
976 GmListElement* next;
977 void* base;
978};
979
980static GmListElement* head = 0;
981static unsigned int gNextAddress = 0;
982static unsigned int gAddressBase = 0;
983static unsigned int gAllocatedSize = 0;
984
985static
986GmListElement* makeGmListElement (void* bas)
987{
988 GmListElement* this;
989 this = (GmListElement*)(void*)LocalAlloc (0, sizeof (GmListElement));
990 assert (this);
991 if (this)
992 {
993 this->base = bas;
994 this->next = head;
995 head = this;
996 }
997 return this;
998}
999
1000void gcleanup ()
1001{
1002 BOOL rval;
1003 assert ( (head == NULL) || (head->base == (void*)gAddressBase));
1004 if (gAddressBase && (gNextAddress - gAddressBase))
1005 {
1006 rval = VirtualFree ((void*)gAddressBase,
1007 gNextAddress - gAddressBase,
1008 MEM_DECOMMIT);
wdenk8bde7f72003-06-27 21:31:46 +00001009 assert (rval);
wdenk217c9da2002-10-25 20:35:49 +00001010 }
1011 while (head)
1012 {
1013 GmListElement* next = head->next;
1014 rval = VirtualFree (head->base, 0, MEM_RELEASE);
1015 assert (rval);
1016 LocalFree (head);
1017 head = next;
1018 }
1019}
1020
1021static
1022void* findRegion (void* start_address, unsigned long size)
1023{
1024 MEMORY_BASIC_INFORMATION info;
1025 if (size >= TOP_MEMORY) return NULL;
1026
1027 while ((unsigned long)start_address + size < TOP_MEMORY)
1028 {
1029 VirtualQuery (start_address, &info, sizeof (info));
1030 if ((info.State == MEM_FREE) && (info.RegionSize >= size))
1031 return start_address;
1032 else
1033 {
wdenk8bde7f72003-06-27 21:31:46 +00001034 /* Requested region is not available so see if the */
1035 /* next region is available. Set 'start_address' */
1036 /* to the next region and call 'VirtualQuery()' */
1037 /* again. */
wdenk217c9da2002-10-25 20:35:49 +00001038
1039 start_address = (char*)info.BaseAddress + info.RegionSize;
1040
wdenk8bde7f72003-06-27 21:31:46 +00001041 /* Make sure we start looking for the next region */
1042 /* on the *next* 64K boundary. Otherwise, even if */
1043 /* the new region is free according to */
1044 /* 'VirtualQuery()', the subsequent call to */
1045 /* 'VirtualAlloc()' (which follows the call to */
1046 /* this routine in 'wsbrk()') will round *down* */
1047 /* the requested address to a 64K boundary which */
1048 /* we already know is an address in the */
1049 /* unavailable region. Thus, the subsequent call */
1050 /* to 'VirtualAlloc()' will fail and bring us back */
1051 /* here, causing us to go into an infinite loop. */
wdenk217c9da2002-10-25 20:35:49 +00001052
1053 start_address =
1054 (void *) AlignPage64K((unsigned long) start_address);
1055 }
1056 }
1057 return NULL;
1058
1059}
1060
1061
1062void* wsbrk (long size)
1063{
1064 void* tmp;
1065 if (size > 0)
1066 {
1067 if (gAddressBase == 0)
1068 {
1069 gAllocatedSize = max (RESERVED_SIZE, AlignPage (size));
1070 gNextAddress = gAddressBase =
1071 (unsigned int)VirtualAlloc (NULL, gAllocatedSize,
1072 MEM_RESERVE, PAGE_NOACCESS);
1073 } else if (AlignPage (gNextAddress + size) > (gAddressBase +
1074gAllocatedSize))
1075 {
1076 long new_size = max (NEXT_SIZE, AlignPage (size));
1077 void* new_address = (void*)(gAddressBase+gAllocatedSize);
1078 do
1079 {
1080 new_address = findRegion (new_address, new_size);
1081
1082 if (new_address == 0)
1083 return (void*)-1;
1084
1085 gAddressBase = gNextAddress =
1086 (unsigned int)VirtualAlloc (new_address, new_size,
1087 MEM_RESERVE, PAGE_NOACCESS);
wdenk8bde7f72003-06-27 21:31:46 +00001088 /* repeat in case of race condition */
1089 /* The region that we found has been snagged */
1090 /* by another thread */
wdenk217c9da2002-10-25 20:35:49 +00001091 }
1092 while (gAddressBase == 0);
1093
1094 assert (new_address == (void*)gAddressBase);
1095
1096 gAllocatedSize = new_size;
1097
1098 if (!makeGmListElement ((void*)gAddressBase))
1099 return (void*)-1;
1100 }
1101 if ((size + gNextAddress) > AlignPage (gNextAddress))
1102 {
1103 void* res;
1104 res = VirtualAlloc ((void*)AlignPage (gNextAddress),
1105 (size + gNextAddress -
1106 AlignPage (gNextAddress)),
1107 MEM_COMMIT, PAGE_READWRITE);
1108 if (res == 0)
1109 return (void*)-1;
1110 }
1111 tmp = (void*)gNextAddress;
1112 gNextAddress = (unsigned int)tmp + size;
1113 return tmp;
1114 }
1115 else if (size < 0)
1116 {
1117 unsigned int alignedGoal = AlignPage (gNextAddress + size);
1118 /* Trim by releasing the virtual memory */
1119 if (alignedGoal >= gAddressBase)
1120 {
1121 VirtualFree ((void*)alignedGoal, gNextAddress - alignedGoal,
1122 MEM_DECOMMIT);
1123 gNextAddress = gNextAddress + size;
1124 return (void*)gNextAddress;
1125 }
1126 else
1127 {
1128 VirtualFree ((void*)gAddressBase, gNextAddress - gAddressBase,
1129 MEM_DECOMMIT);
1130 gNextAddress = gAddressBase;
1131 return (void*)-1;
1132 }
1133 }
1134 else
1135 {
1136 return (void*)gNextAddress;
1137 }
1138}
1139
1140#endif
1141
1142
1143
1144/*
1145 Type declarations
1146*/
1147
1148
1149struct malloc_chunk
1150{
1151 INTERNAL_SIZE_T prev_size; /* Size of previous chunk (if free). */
1152 INTERNAL_SIZE_T size; /* Size in bytes, including overhead. */
1153 struct malloc_chunk* fd; /* double links -- used only if free. */
1154 struct malloc_chunk* bk;
Joakim Tjernlund1ba91ba2010-10-14 08:51:34 +02001155} __attribute__((__may_alias__)) ;
wdenk217c9da2002-10-25 20:35:49 +00001156
1157typedef struct malloc_chunk* mchunkptr;
1158
1159/*
1160
1161 malloc_chunk details:
1162
1163 (The following includes lightly edited explanations by Colin Plumb.)
1164
1165 Chunks of memory are maintained using a `boundary tag' method as
1166 described in e.g., Knuth or Standish. (See the paper by Paul
1167 Wilson ftp://ftp.cs.utexas.edu/pub/garbage/allocsrv.ps for a
1168 survey of such techniques.) Sizes of free chunks are stored both
1169 in the front of each chunk and at the end. This makes
1170 consolidating fragmented chunks into bigger chunks very fast. The
1171 size fields also hold bits representing whether chunks are free or
1172 in use.
1173
1174 An allocated chunk looks like this:
1175
1176
1177 chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk8bde7f72003-06-27 21:31:46 +00001178 | Size of previous chunk, if allocated | |
1179 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1180 | Size of chunk, in bytes |P|
wdenk217c9da2002-10-25 20:35:49 +00001181 mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk8bde7f72003-06-27 21:31:46 +00001182 | User data starts here... .
1183 . .
1184 . (malloc_usable_space() bytes) .
1185 . |
wdenk217c9da2002-10-25 20:35:49 +00001186nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk8bde7f72003-06-27 21:31:46 +00001187 | Size of chunk |
1188 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk217c9da2002-10-25 20:35:49 +00001189
1190
1191 Where "chunk" is the front of the chunk for the purpose of most of
1192 the malloc code, but "mem" is the pointer that is returned to the
1193 user. "Nextchunk" is the beginning of the next contiguous chunk.
1194
1195 Chunks always begin on even word boundries, so the mem portion
1196 (which is returned to the user) is also on an even word boundary, and
1197 thus double-word aligned.
1198
1199 Free chunks are stored in circular doubly-linked lists, and look like this:
1200
1201 chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk8bde7f72003-06-27 21:31:46 +00001202 | Size of previous chunk |
1203 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk217c9da2002-10-25 20:35:49 +00001204 `head:' | Size of chunk, in bytes |P|
1205 mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk8bde7f72003-06-27 21:31:46 +00001206 | Forward pointer to next chunk in list |
1207 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1208 | Back pointer to previous chunk in list |
1209 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1210 | Unused space (may be 0 bytes long) .
1211 . .
1212 . |
wdenk217c9da2002-10-25 20:35:49 +00001213nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1214 `foot:' | Size of chunk, in bytes |
wdenk8bde7f72003-06-27 21:31:46 +00001215 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
wdenk217c9da2002-10-25 20:35:49 +00001216
1217 The P (PREV_INUSE) bit, stored in the unused low-order bit of the
1218 chunk size (which is always a multiple of two words), is an in-use
1219 bit for the *previous* chunk. If that bit is *clear*, then the
1220 word before the current chunk size contains the previous chunk
1221 size, and can be used to find the front of the previous chunk.
1222 (The very first chunk allocated always has this bit set,
1223 preventing access to non-existent (or non-owned) memory.)
1224
1225 Note that the `foot' of the current chunk is actually represented
1226 as the prev_size of the NEXT chunk. (This makes it easier to
1227 deal with alignments etc).
1228
1229 The two exceptions to all this are
1230
1231 1. The special chunk `top', which doesn't bother using the
wdenk8bde7f72003-06-27 21:31:46 +00001232 trailing size field since there is no
1233 next contiguous chunk that would have to index off it. (After
1234 initialization, `top' is forced to always exist. If it would
1235 become less than MINSIZE bytes long, it is replenished via
1236 malloc_extend_top.)
wdenk217c9da2002-10-25 20:35:49 +00001237
1238 2. Chunks allocated via mmap, which have the second-lowest-order
wdenk8bde7f72003-06-27 21:31:46 +00001239 bit (IS_MMAPPED) set in their size fields. Because they are
1240 never merged or traversed from any other chunk, they have no
1241 foot size or inuse information.
wdenk217c9da2002-10-25 20:35:49 +00001242
1243 Available chunks are kept in any of several places (all declared below):
1244
1245 * `av': An array of chunks serving as bin headers for consolidated
1246 chunks. Each bin is doubly linked. The bins are approximately
1247 proportionally (log) spaced. There are a lot of these bins
1248 (128). This may look excessive, but works very well in
1249 practice. All procedures maintain the invariant that no
1250 consolidated chunk physically borders another one. Chunks in
1251 bins are kept in size order, with ties going to the
1252 approximately least recently used chunk.
1253
1254 The chunks in each bin are maintained in decreasing sorted order by
1255 size. This is irrelevant for the small bins, which all contain
1256 the same-sized chunks, but facilitates best-fit allocation for
1257 larger chunks. (These lists are just sequential. Keeping them in
1258 order almost never requires enough traversal to warrant using
1259 fancier ordered data structures.) Chunks of the same size are
1260 linked with the most recently freed at the front, and allocations
1261 are taken from the back. This results in LRU or FIFO allocation
1262 order, which tends to give each chunk an equal opportunity to be
1263 consolidated with adjacent freed chunks, resulting in larger free
1264 chunks and less fragmentation.
1265
1266 * `top': The top-most available chunk (i.e., the one bordering the
1267 end of available memory) is treated specially. It is never
1268 included in any bin, is used only if no other chunk is
1269 available, and is released back to the system if it is very
1270 large (see M_TRIM_THRESHOLD).
1271
1272 * `last_remainder': A bin holding only the remainder of the
1273 most recently split (non-top) chunk. This bin is checked
1274 before other non-fitting chunks, so as to provide better
1275 locality for runs of sequentially allocated chunks.
1276
1277 * Implicitly, through the host system's memory mapping tables.
1278 If supported, requests greater than a threshold are usually
1279 serviced via calls to mmap, and then later released via munmap.
1280
1281*/
wdenk217c9da2002-10-25 20:35:49 +00001282
wdenk217c9da2002-10-25 20:35:49 +00001283/* sizes, alignments */
1284
1285#define SIZE_SZ (sizeof(INTERNAL_SIZE_T))
1286#define MALLOC_ALIGNMENT (SIZE_SZ + SIZE_SZ)
1287#define MALLOC_ALIGN_MASK (MALLOC_ALIGNMENT - 1)
1288#define MINSIZE (sizeof(struct malloc_chunk))
1289
1290/* conversion from malloc headers to user pointers, and back */
1291
1292#define chunk2mem(p) ((Void_t*)((char*)(p) + 2*SIZE_SZ))
1293#define mem2chunk(mem) ((mchunkptr)((char*)(mem) - 2*SIZE_SZ))
1294
1295/* pad request bytes into a usable size */
1296
1297#define request2size(req) \
1298 (((long)((req) + (SIZE_SZ + MALLOC_ALIGN_MASK)) < \
1299 (long)(MINSIZE + MALLOC_ALIGN_MASK)) ? MINSIZE : \
1300 (((req) + (SIZE_SZ + MALLOC_ALIGN_MASK)) & ~(MALLOC_ALIGN_MASK)))
1301
1302/* Check if m has acceptable alignment */
1303
1304#define aligned_OK(m) (((unsigned long)((m)) & (MALLOC_ALIGN_MASK)) == 0)
1305
1306
1307
1308
1309/*
1310 Physical chunk operations
1311*/
1312
1313
1314/* size field is or'ed with PREV_INUSE when previous adjacent chunk in use */
1315
1316#define PREV_INUSE 0x1
1317
1318/* size field is or'ed with IS_MMAPPED if the chunk was obtained with mmap() */
1319
1320#define IS_MMAPPED 0x2
1321
1322/* Bits to mask off when extracting size */
1323
1324#define SIZE_BITS (PREV_INUSE|IS_MMAPPED)
1325
1326
1327/* Ptr to next physical malloc_chunk. */
1328
1329#define next_chunk(p) ((mchunkptr)( ((char*)(p)) + ((p)->size & ~PREV_INUSE) ))
1330
1331/* Ptr to previous physical malloc_chunk */
1332
1333#define prev_chunk(p)\
1334 ((mchunkptr)( ((char*)(p)) - ((p)->prev_size) ))
1335
1336
1337/* Treat space at ptr + offset as a chunk */
1338
1339#define chunk_at_offset(p, s) ((mchunkptr)(((char*)(p)) + (s)))
1340
1341
1342
1343
1344/*
1345 Dealing with use bits
1346*/
1347
1348/* extract p's inuse bit */
1349
1350#define inuse(p)\
1351((((mchunkptr)(((char*)(p))+((p)->size & ~PREV_INUSE)))->size) & PREV_INUSE)
1352
1353/* extract inuse bit of previous chunk */
1354
1355#define prev_inuse(p) ((p)->size & PREV_INUSE)
1356
1357/* check for mmap()'ed chunk */
1358
1359#define chunk_is_mmapped(p) ((p)->size & IS_MMAPPED)
1360
1361/* set/clear chunk as in use without otherwise disturbing */
1362
1363#define set_inuse(p)\
1364((mchunkptr)(((char*)(p)) + ((p)->size & ~PREV_INUSE)))->size |= PREV_INUSE
1365
1366#define clear_inuse(p)\
1367((mchunkptr)(((char*)(p)) + ((p)->size & ~PREV_INUSE)))->size &= ~(PREV_INUSE)
1368
1369/* check/set/clear inuse bits in known places */
1370
1371#define inuse_bit_at_offset(p, s)\
1372 (((mchunkptr)(((char*)(p)) + (s)))->size & PREV_INUSE)
1373
1374#define set_inuse_bit_at_offset(p, s)\
1375 (((mchunkptr)(((char*)(p)) + (s)))->size |= PREV_INUSE)
1376
1377#define clear_inuse_bit_at_offset(p, s)\
1378 (((mchunkptr)(((char*)(p)) + (s)))->size &= ~(PREV_INUSE))
1379
1380
1381
1382
1383/*
1384 Dealing with size fields
1385*/
1386
1387/* Get size, ignoring use bits */
1388
1389#define chunksize(p) ((p)->size & ~(SIZE_BITS))
1390
1391/* Set size at head, without disturbing its use bit */
1392
1393#define set_head_size(p, s) ((p)->size = (((p)->size & PREV_INUSE) | (s)))
1394
1395/* Set size/use ignoring previous bits in header */
1396
1397#define set_head(p, s) ((p)->size = (s))
1398
1399/* Set size at footer (only when chunk is not in use) */
1400
1401#define set_foot(p, s) (((mchunkptr)((char*)(p) + (s)))->prev_size = (s))
1402
1403
1404
1405
1406
1407/*
1408 Bins
1409
1410 The bins, `av_' are an array of pairs of pointers serving as the
1411 heads of (initially empty) doubly-linked lists of chunks, laid out
1412 in a way so that each pair can be treated as if it were in a
1413 malloc_chunk. (This way, the fd/bk offsets for linking bin heads
1414 and chunks are the same).
1415
1416 Bins for sizes < 512 bytes contain chunks of all the same size, spaced
1417 8 bytes apart. Larger bins are approximately logarithmically
1418 spaced. (See the table below.) The `av_' array is never mentioned
1419 directly in the code, but instead via bin access macros.
1420
1421 Bin layout:
1422
1423 64 bins of size 8
1424 32 bins of size 64
1425 16 bins of size 512
1426 8 bins of size 4096
1427 4 bins of size 32768
1428 2 bins of size 262144
1429 1 bin of size what's left
1430
1431 There is actually a little bit of slop in the numbers in bin_index
1432 for the sake of speed. This makes no difference elsewhere.
1433
1434 The special chunks `top' and `last_remainder' get their own bins,
1435 (this is implemented via yet more trickery with the av_ array),
1436 although `top' is never properly linked to its bin since it is
1437 always handled specially.
1438
1439*/
1440
1441#define NAV 128 /* number of bins */
1442
1443typedef struct malloc_chunk* mbinptr;
1444
1445/* access macros */
1446
1447#define bin_at(i) ((mbinptr)((char*)&(av_[2*(i) + 2]) - 2*SIZE_SZ))
1448#define next_bin(b) ((mbinptr)((char*)(b) + 2 * sizeof(mbinptr)))
1449#define prev_bin(b) ((mbinptr)((char*)(b) - 2 * sizeof(mbinptr)))
1450
1451/*
1452 The first 2 bins are never indexed. The corresponding av_ cells are instead
1453 used for bookkeeping. This is not to save space, but to simplify
1454 indexing, maintain locality, and avoid some initialization tests.
1455*/
1456
Stefan Roesef2302d42008-08-06 14:05:38 +02001457#define top (av_[2]) /* The topmost chunk */
wdenk217c9da2002-10-25 20:35:49 +00001458#define last_remainder (bin_at(1)) /* remainder from last split */
1459
1460
1461/*
1462 Because top initially points to its own bin with initial
1463 zero size, thus forcing extension on the first malloc request,
1464 we avoid having any special code in malloc to check whether
1465 it even exists yet. But we still need to in malloc_extend_top.
1466*/
1467
1468#define initial_top ((mchunkptr)(bin_at(0)))
1469
1470/* Helper macro to initialize bins */
1471
1472#define IAV(i) bin_at(i), bin_at(i)
1473
1474static mbinptr av_[NAV * 2 + 2] = {
1475 0, 0,
1476 IAV(0), IAV(1), IAV(2), IAV(3), IAV(4), IAV(5), IAV(6), IAV(7),
1477 IAV(8), IAV(9), IAV(10), IAV(11), IAV(12), IAV(13), IAV(14), IAV(15),
1478 IAV(16), IAV(17), IAV(18), IAV(19), IAV(20), IAV(21), IAV(22), IAV(23),
1479 IAV(24), IAV(25), IAV(26), IAV(27), IAV(28), IAV(29), IAV(30), IAV(31),
1480 IAV(32), IAV(33), IAV(34), IAV(35), IAV(36), IAV(37), IAV(38), IAV(39),
1481 IAV(40), IAV(41), IAV(42), IAV(43), IAV(44), IAV(45), IAV(46), IAV(47),
1482 IAV(48), IAV(49), IAV(50), IAV(51), IAV(52), IAV(53), IAV(54), IAV(55),
1483 IAV(56), IAV(57), IAV(58), IAV(59), IAV(60), IAV(61), IAV(62), IAV(63),
1484 IAV(64), IAV(65), IAV(66), IAV(67), IAV(68), IAV(69), IAV(70), IAV(71),
1485 IAV(72), IAV(73), IAV(74), IAV(75), IAV(76), IAV(77), IAV(78), IAV(79),
1486 IAV(80), IAV(81), IAV(82), IAV(83), IAV(84), IAV(85), IAV(86), IAV(87),
1487 IAV(88), IAV(89), IAV(90), IAV(91), IAV(92), IAV(93), IAV(94), IAV(95),
1488 IAV(96), IAV(97), IAV(98), IAV(99), IAV(100), IAV(101), IAV(102), IAV(103),
1489 IAV(104), IAV(105), IAV(106), IAV(107), IAV(108), IAV(109), IAV(110), IAV(111),
1490 IAV(112), IAV(113), IAV(114), IAV(115), IAV(116), IAV(117), IAV(118), IAV(119),
1491 IAV(120), IAV(121), IAV(122), IAV(123), IAV(124), IAV(125), IAV(126), IAV(127)
1492};
1493
Wolfgang Denk2e5167c2010-10-28 20:00:11 +02001494#ifdef CONFIG_NEEDS_MANUAL_RELOC
wdenk217c9da2002-10-25 20:35:49 +00001495void malloc_bin_reloc (void)
1496{
wdenk217c9da2002-10-25 20:35:49 +00001497 unsigned long *p = (unsigned long *)(&av_[2]);
1498 int i;
1499 for (i=2; i<(sizeof(av_)/sizeof(mbinptr)); ++i) {
1500 *p++ += gd->reloc_off;
1501 }
1502}
Peter Tyser521af042009-09-21 11:20:36 -05001503#endif
Peter Tyser5e93bd12009-08-21 23:05:19 -05001504
1505ulong mem_malloc_start = 0;
1506ulong mem_malloc_end = 0;
1507ulong mem_malloc_brk = 0;
1508
1509void *sbrk(ptrdiff_t increment)
1510{
1511 ulong old = mem_malloc_brk;
1512 ulong new = old + increment;
1513
Kumar Gala6163f5b2010-11-15 18:41:43 -06001514 /*
1515 * if we are giving memory back make sure we clear it out since
1516 * we set MORECORE_CLEARS to 1
1517 */
1518 if (increment < 0)
1519 memset((void *)new, 0, -increment);
1520
Peter Tyser5e93bd12009-08-21 23:05:19 -05001521 if ((new < mem_malloc_start) || (new > mem_malloc_end))
karl.beldan@gmail.comae30b8c2010-04-06 22:18:08 +02001522 return (void *)MORECORE_FAILURE;
Peter Tyser5e93bd12009-08-21 23:05:19 -05001523
1524 mem_malloc_brk = new;
1525
1526 return (void *)old;
1527}
wdenk217c9da2002-10-25 20:35:49 +00001528
Peter Tyserd4e8ada2009-08-21 23:05:21 -05001529void mem_malloc_init(ulong start, ulong size)
1530{
1531 mem_malloc_start = start;
1532 mem_malloc_end = start + size;
1533 mem_malloc_brk = start;
1534
1535 memset((void *)mem_malloc_start, 0, size);
1536}
Peter Tyserd4e8ada2009-08-21 23:05:21 -05001537
wdenk217c9da2002-10-25 20:35:49 +00001538/* field-extraction macros */
1539
1540#define first(b) ((b)->fd)
1541#define last(b) ((b)->bk)
1542
1543/*
1544 Indexing into bins
1545*/
1546
1547#define bin_index(sz) \
1548(((((unsigned long)(sz)) >> 9) == 0) ? (((unsigned long)(sz)) >> 3): \
1549 ((((unsigned long)(sz)) >> 9) <= 4) ? 56 + (((unsigned long)(sz)) >> 6): \
1550 ((((unsigned long)(sz)) >> 9) <= 20) ? 91 + (((unsigned long)(sz)) >> 9): \
1551 ((((unsigned long)(sz)) >> 9) <= 84) ? 110 + (((unsigned long)(sz)) >> 12): \
1552 ((((unsigned long)(sz)) >> 9) <= 340) ? 119 + (((unsigned long)(sz)) >> 15): \
1553 ((((unsigned long)(sz)) >> 9) <= 1364) ? 124 + (((unsigned long)(sz)) >> 18): \
wdenk8bde7f72003-06-27 21:31:46 +00001554 126)
wdenk217c9da2002-10-25 20:35:49 +00001555/*
1556 bins for chunks < 512 are all spaced 8 bytes apart, and hold
1557 identically sized chunks. This is exploited in malloc.
1558*/
1559
1560#define MAX_SMALLBIN 63
1561#define MAX_SMALLBIN_SIZE 512
1562#define SMALLBIN_WIDTH 8
1563
1564#define smallbin_index(sz) (((unsigned long)(sz)) >> 3)
1565
1566/*
1567 Requests are `small' if both the corresponding and the next bin are small
1568*/
1569
1570#define is_small_request(nb) (nb < MAX_SMALLBIN_SIZE - SMALLBIN_WIDTH)
1571
1572
1573
1574/*
1575 To help compensate for the large number of bins, a one-level index
1576 structure is used for bin-by-bin searching. `binblocks' is a
1577 one-word bitvector recording whether groups of BINBLOCKWIDTH bins
1578 have any (possibly) non-empty bins, so they can be skipped over
1579 all at once during during traversals. The bits are NOT always
1580 cleared as soon as all bins in a block are empty, but instead only
1581 when all are noticed to be empty during traversal in malloc.
1582*/
1583
1584#define BINBLOCKWIDTH 4 /* bins per block */
1585
Stefan Roesef2302d42008-08-06 14:05:38 +02001586#define binblocks_r ((INTERNAL_SIZE_T)av_[1]) /* bitvector of nonempty blocks */
1587#define binblocks_w (av_[1])
wdenk217c9da2002-10-25 20:35:49 +00001588
1589/* bin<->block macros */
1590
1591#define idx2binblock(ix) ((unsigned)1 << (ix / BINBLOCKWIDTH))
Stefan Roesef2302d42008-08-06 14:05:38 +02001592#define mark_binblock(ii) (binblocks_w = (mbinptr)(binblocks_r | idx2binblock(ii)))
1593#define clear_binblock(ii) (binblocks_w = (mbinptr)(binblocks_r & ~(idx2binblock(ii))))
wdenk217c9da2002-10-25 20:35:49 +00001594
1595
1596
1597
1598
1599/* Other static bookkeeping data */
1600
1601/* variables holding tunable values */
1602
1603static unsigned long trim_threshold = DEFAULT_TRIM_THRESHOLD;
1604static unsigned long top_pad = DEFAULT_TOP_PAD;
1605static unsigned int n_mmaps_max = DEFAULT_MMAP_MAX;
1606static unsigned long mmap_threshold = DEFAULT_MMAP_THRESHOLD;
1607
1608/* The first value returned from sbrk */
1609static char* sbrk_base = (char*)(-1);
1610
1611/* The maximum memory obtained from system via sbrk */
1612static unsigned long max_sbrked_mem = 0;
1613
1614/* The maximum via either sbrk or mmap */
1615static unsigned long max_total_mem = 0;
1616
1617/* internal working copy of mallinfo */
1618static struct mallinfo current_mallinfo = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
1619
1620/* The total memory obtained from system via sbrk */
1621#define sbrked_mem (current_mallinfo.arena)
1622
1623/* Tracking mmaps */
1624
Wolfgang Denkea882ba2010-06-20 23:33:59 +02001625#ifdef DEBUG
wdenk217c9da2002-10-25 20:35:49 +00001626static unsigned int n_mmaps = 0;
Wolfgang Denkea882ba2010-06-20 23:33:59 +02001627#endif /* DEBUG */
wdenk217c9da2002-10-25 20:35:49 +00001628static unsigned long mmapped_mem = 0;
1629#if HAVE_MMAP
1630static unsigned int max_n_mmaps = 0;
1631static unsigned long max_mmapped_mem = 0;
1632#endif
1633
1634
1635
1636/*
1637 Debugging support
1638*/
1639
1640#ifdef DEBUG
1641
1642
1643/*
1644 These routines make a number of assertions about the states
1645 of data structures that should be true at all times. If any
1646 are not true, it's very likely that a user program has somehow
1647 trashed memory. (It's also possible that there is a coding error
1648 in malloc. In which case, please report it!)
1649*/
1650
1651#if __STD_C
1652static void do_check_chunk(mchunkptr p)
1653#else
1654static void do_check_chunk(p) mchunkptr p;
1655#endif
1656{
1657#if 0 /* causes warnings because assert() is off */
1658 INTERNAL_SIZE_T sz = p->size & ~PREV_INUSE;
1659#endif /* 0 */
1660
1661 /* No checkable chunk is mmapped */
1662 assert(!chunk_is_mmapped(p));
1663
1664 /* Check for legal address ... */
1665 assert((char*)p >= sbrk_base);
1666 if (p != top)
1667 assert((char*)p + sz <= (char*)top);
1668 else
1669 assert((char*)p + sz <= sbrk_base + sbrked_mem);
1670
1671}
1672
1673
1674#if __STD_C
1675static void do_check_free_chunk(mchunkptr p)
1676#else
1677static void do_check_free_chunk(p) mchunkptr p;
1678#endif
1679{
1680 INTERNAL_SIZE_T sz = p->size & ~PREV_INUSE;
1681#if 0 /* causes warnings because assert() is off */
1682 mchunkptr next = chunk_at_offset(p, sz);
1683#endif /* 0 */
1684
1685 do_check_chunk(p);
1686
1687 /* Check whether it claims to be free ... */
1688 assert(!inuse(p));
1689
1690 /* Unless a special marker, must have OK fields */
1691 if ((long)sz >= (long)MINSIZE)
1692 {
1693 assert((sz & MALLOC_ALIGN_MASK) == 0);
1694 assert(aligned_OK(chunk2mem(p)));
1695 /* ... matching footer field */
1696 assert(next->prev_size == sz);
1697 /* ... and is fully consolidated */
1698 assert(prev_inuse(p));
1699 assert (next == top || inuse(next));
1700
1701 /* ... and has minimally sane links */
1702 assert(p->fd->bk == p);
1703 assert(p->bk->fd == p);
1704 }
1705 else /* markers are always of size SIZE_SZ */
1706 assert(sz == SIZE_SZ);
1707}
1708
1709#if __STD_C
1710static void do_check_inuse_chunk(mchunkptr p)
1711#else
1712static void do_check_inuse_chunk(p) mchunkptr p;
1713#endif
1714{
1715 mchunkptr next = next_chunk(p);
1716 do_check_chunk(p);
1717
1718 /* Check whether it claims to be in use ... */
1719 assert(inuse(p));
1720
1721 /* ... and is surrounded by OK chunks.
1722 Since more things can be checked with free chunks than inuse ones,
1723 if an inuse chunk borders them and debug is on, it's worth doing them.
1724 */
1725 if (!prev_inuse(p))
1726 {
1727 mchunkptr prv = prev_chunk(p);
1728 assert(next_chunk(prv) == p);
1729 do_check_free_chunk(prv);
1730 }
1731 if (next == top)
1732 {
1733 assert(prev_inuse(next));
1734 assert(chunksize(next) >= MINSIZE);
1735 }
1736 else if (!inuse(next))
1737 do_check_free_chunk(next);
1738
1739}
1740
1741#if __STD_C
1742static void do_check_malloced_chunk(mchunkptr p, INTERNAL_SIZE_T s)
1743#else
1744static void do_check_malloced_chunk(p, s) mchunkptr p; INTERNAL_SIZE_T s;
1745#endif
1746{
1747#if 0 /* causes warnings because assert() is off */
1748 INTERNAL_SIZE_T sz = p->size & ~PREV_INUSE;
1749 long room = sz - s;
1750#endif /* 0 */
1751
1752 do_check_inuse_chunk(p);
1753
1754 /* Legal size ... */
1755 assert((long)sz >= (long)MINSIZE);
1756 assert((sz & MALLOC_ALIGN_MASK) == 0);
1757 assert(room >= 0);
1758 assert(room < (long)MINSIZE);
1759
1760 /* ... and alignment */
1761 assert(aligned_OK(chunk2mem(p)));
1762
1763
1764 /* ... and was allocated at front of an available chunk */
1765 assert(prev_inuse(p));
1766
1767}
1768
1769
1770#define check_free_chunk(P) do_check_free_chunk(P)
1771#define check_inuse_chunk(P) do_check_inuse_chunk(P)
1772#define check_chunk(P) do_check_chunk(P)
1773#define check_malloced_chunk(P,N) do_check_malloced_chunk(P,N)
1774#else
1775#define check_free_chunk(P)
1776#define check_inuse_chunk(P)
1777#define check_chunk(P)
1778#define check_malloced_chunk(P,N)
1779#endif
1780
1781
1782
1783/*
1784 Macro-based internal utilities
1785*/
1786
1787
1788/*
1789 Linking chunks in bin lists.
1790 Call these only with variables, not arbitrary expressions, as arguments.
1791*/
1792
1793/*
1794 Place chunk p of size s in its bin, in size order,
1795 putting it ahead of others of same size.
1796*/
1797
1798
1799#define frontlink(P, S, IDX, BK, FD) \
1800{ \
1801 if (S < MAX_SMALLBIN_SIZE) \
1802 { \
1803 IDX = smallbin_index(S); \
1804 mark_binblock(IDX); \
1805 BK = bin_at(IDX); \
1806 FD = BK->fd; \
1807 P->bk = BK; \
1808 P->fd = FD; \
1809 FD->bk = BK->fd = P; \
1810 } \
1811 else \
1812 { \
1813 IDX = bin_index(S); \
1814 BK = bin_at(IDX); \
1815 FD = BK->fd; \
1816 if (FD == BK) mark_binblock(IDX); \
1817 else \
1818 { \
1819 while (FD != BK && S < chunksize(FD)) FD = FD->fd; \
1820 BK = FD->bk; \
1821 } \
1822 P->bk = BK; \
1823 P->fd = FD; \
1824 FD->bk = BK->fd = P; \
1825 } \
1826}
1827
1828
1829/* take a chunk off a list */
1830
1831#define unlink(P, BK, FD) \
1832{ \
1833 BK = P->bk; \
1834 FD = P->fd; \
1835 FD->bk = BK; \
1836 BK->fd = FD; \
1837} \
1838
1839/* Place p as the last remainder */
1840
1841#define link_last_remainder(P) \
1842{ \
1843 last_remainder->fd = last_remainder->bk = P; \
1844 P->fd = P->bk = last_remainder; \
1845}
1846
1847/* Clear the last_remainder bin */
1848
1849#define clear_last_remainder \
1850 (last_remainder->fd = last_remainder->bk = last_remainder)
1851
1852
wdenk217c9da2002-10-25 20:35:49 +00001853
1854
1855
1856/* Routines dealing with mmap(). */
1857
1858#if HAVE_MMAP
1859
1860#if __STD_C
1861static mchunkptr mmap_chunk(size_t size)
1862#else
1863static mchunkptr mmap_chunk(size) size_t size;
1864#endif
1865{
1866 size_t page_mask = malloc_getpagesize - 1;
1867 mchunkptr p;
1868
1869#ifndef MAP_ANONYMOUS
1870 static int fd = -1;
1871#endif
1872
1873 if(n_mmaps >= n_mmaps_max) return 0; /* too many regions */
1874
1875 /* For mmapped chunks, the overhead is one SIZE_SZ unit larger, because
1876 * there is no following chunk whose prev_size field could be used.
1877 */
1878 size = (size + SIZE_SZ + page_mask) & ~page_mask;
1879
1880#ifdef MAP_ANONYMOUS
1881 p = (mchunkptr)mmap(0, size, PROT_READ|PROT_WRITE,
1882 MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
1883#else /* !MAP_ANONYMOUS */
1884 if (fd < 0)
1885 {
1886 fd = open("/dev/zero", O_RDWR);
1887 if(fd < 0) return 0;
1888 }
1889 p = (mchunkptr)mmap(0, size, PROT_READ|PROT_WRITE, MAP_PRIVATE, fd, 0);
1890#endif
1891
1892 if(p == (mchunkptr)-1) return 0;
1893
1894 n_mmaps++;
1895 if (n_mmaps > max_n_mmaps) max_n_mmaps = n_mmaps;
1896
1897 /* We demand that eight bytes into a page must be 8-byte aligned. */
1898 assert(aligned_OK(chunk2mem(p)));
1899
1900 /* The offset to the start of the mmapped region is stored
1901 * in the prev_size field of the chunk; normally it is zero,
1902 * but that can be changed in memalign().
1903 */
1904 p->prev_size = 0;
1905 set_head(p, size|IS_MMAPPED);
1906
1907 mmapped_mem += size;
1908 if ((unsigned long)mmapped_mem > (unsigned long)max_mmapped_mem)
1909 max_mmapped_mem = mmapped_mem;
1910 if ((unsigned long)(mmapped_mem + sbrked_mem) > (unsigned long)max_total_mem)
1911 max_total_mem = mmapped_mem + sbrked_mem;
1912 return p;
1913}
1914
1915#if __STD_C
1916static void munmap_chunk(mchunkptr p)
1917#else
1918static void munmap_chunk(p) mchunkptr p;
1919#endif
1920{
1921 INTERNAL_SIZE_T size = chunksize(p);
1922 int ret;
1923
1924 assert (chunk_is_mmapped(p));
1925 assert(! ((char*)p >= sbrk_base && (char*)p < sbrk_base + sbrked_mem));
1926 assert((n_mmaps > 0));
1927 assert(((p->prev_size + size) & (malloc_getpagesize-1)) == 0);
1928
1929 n_mmaps--;
1930 mmapped_mem -= (size + p->prev_size);
1931
1932 ret = munmap((char *)p - p->prev_size, size + p->prev_size);
1933
1934 /* munmap returns non-zero on failure */
1935 assert(ret == 0);
1936}
1937
1938#if HAVE_MREMAP
1939
1940#if __STD_C
1941static mchunkptr mremap_chunk(mchunkptr p, size_t new_size)
1942#else
1943static mchunkptr mremap_chunk(p, new_size) mchunkptr p; size_t new_size;
1944#endif
1945{
1946 size_t page_mask = malloc_getpagesize - 1;
1947 INTERNAL_SIZE_T offset = p->prev_size;
1948 INTERNAL_SIZE_T size = chunksize(p);
1949 char *cp;
1950
1951 assert (chunk_is_mmapped(p));
1952 assert(! ((char*)p >= sbrk_base && (char*)p < sbrk_base + sbrked_mem));
1953 assert((n_mmaps > 0));
1954 assert(((size + offset) & (malloc_getpagesize-1)) == 0);
1955
1956 /* Note the extra SIZE_SZ overhead as in mmap_chunk(). */
1957 new_size = (new_size + offset + SIZE_SZ + page_mask) & ~page_mask;
1958
1959 cp = (char *)mremap((char *)p - offset, size + offset, new_size, 1);
1960
1961 if (cp == (char *)-1) return 0;
1962
1963 p = (mchunkptr)(cp + offset);
1964
1965 assert(aligned_OK(chunk2mem(p)));
1966
1967 assert((p->prev_size == offset));
1968 set_head(p, (new_size - offset)|IS_MMAPPED);
1969
1970 mmapped_mem -= size + offset;
1971 mmapped_mem += new_size;
1972 if ((unsigned long)mmapped_mem > (unsigned long)max_mmapped_mem)
1973 max_mmapped_mem = mmapped_mem;
1974 if ((unsigned long)(mmapped_mem + sbrked_mem) > (unsigned long)max_total_mem)
1975 max_total_mem = mmapped_mem + sbrked_mem;
1976 return p;
1977}
1978
1979#endif /* HAVE_MREMAP */
1980
1981#endif /* HAVE_MMAP */
1982
1983
1984
1985
1986/*
1987 Extend the top-most chunk by obtaining memory from system.
1988 Main interface to sbrk (but see also malloc_trim).
1989*/
1990
1991#if __STD_C
1992static void malloc_extend_top(INTERNAL_SIZE_T nb)
1993#else
1994static void malloc_extend_top(nb) INTERNAL_SIZE_T nb;
1995#endif
1996{
1997 char* brk; /* return value from sbrk */
1998 INTERNAL_SIZE_T front_misalign; /* unusable bytes at front of sbrked space */
1999 INTERNAL_SIZE_T correction; /* bytes for 2nd sbrk call */
2000 char* new_brk; /* return of 2nd sbrk call */
2001 INTERNAL_SIZE_T top_size; /* new size of top chunk */
2002
2003 mchunkptr old_top = top; /* Record state of old top */
2004 INTERNAL_SIZE_T old_top_size = chunksize(old_top);
2005 char* old_end = (char*)(chunk_at_offset(old_top, old_top_size));
2006
2007 /* Pad request with top_pad plus minimal overhead */
2008
2009 INTERNAL_SIZE_T sbrk_size = nb + top_pad + MINSIZE;
2010 unsigned long pagesz = malloc_getpagesize;
2011
2012 /* If not the first time through, round to preserve page boundary */
2013 /* Otherwise, we need to correct to a page size below anyway. */
2014 /* (We also correct below if an intervening foreign sbrk call.) */
2015
2016 if (sbrk_base != (char*)(-1))
2017 sbrk_size = (sbrk_size + (pagesz - 1)) & ~(pagesz - 1);
2018
2019 brk = (char*)(MORECORE (sbrk_size));
2020
2021 /* Fail if sbrk failed or if a foreign sbrk call killed our space */
2022 if (brk == (char*)(MORECORE_FAILURE) ||
2023 (brk < old_end && old_top != initial_top))
2024 return;
2025
2026 sbrked_mem += sbrk_size;
2027
2028 if (brk == old_end) /* can just add bytes to current top */
2029 {
2030 top_size = sbrk_size + old_top_size;
2031 set_head(top, top_size | PREV_INUSE);
2032 }
2033 else
2034 {
2035 if (sbrk_base == (char*)(-1)) /* First time through. Record base */
2036 sbrk_base = brk;
2037 else /* Someone else called sbrk(). Count those bytes as sbrked_mem. */
2038 sbrked_mem += brk - (char*)old_end;
2039
2040 /* Guarantee alignment of first new chunk made from this space */
2041 front_misalign = (unsigned long)chunk2mem(brk) & MALLOC_ALIGN_MASK;
2042 if (front_misalign > 0)
2043 {
2044 correction = (MALLOC_ALIGNMENT) - front_misalign;
2045 brk += correction;
2046 }
2047 else
2048 correction = 0;
2049
2050 /* Guarantee the next brk will be at a page boundary */
2051
2052 correction += ((((unsigned long)(brk + sbrk_size))+(pagesz-1)) &
wdenk8bde7f72003-06-27 21:31:46 +00002053 ~(pagesz - 1)) - ((unsigned long)(brk + sbrk_size));
wdenk217c9da2002-10-25 20:35:49 +00002054
2055 /* Allocate correction */
2056 new_brk = (char*)(MORECORE (correction));
2057 if (new_brk == (char*)(MORECORE_FAILURE)) return;
2058
2059 sbrked_mem += correction;
2060
2061 top = (mchunkptr)brk;
2062 top_size = new_brk - brk + correction;
2063 set_head(top, top_size | PREV_INUSE);
2064
2065 if (old_top != initial_top)
2066 {
2067
2068 /* There must have been an intervening foreign sbrk call. */
2069 /* A double fencepost is necessary to prevent consolidation */
2070
2071 /* If not enough space to do this, then user did something very wrong */
2072 if (old_top_size < MINSIZE)
2073 {
wdenk8bde7f72003-06-27 21:31:46 +00002074 set_head(top, PREV_INUSE); /* will force null return from malloc */
2075 return;
wdenk217c9da2002-10-25 20:35:49 +00002076 }
2077
2078 /* Also keep size a multiple of MALLOC_ALIGNMENT */
2079 old_top_size = (old_top_size - 3*SIZE_SZ) & ~MALLOC_ALIGN_MASK;
2080 set_head_size(old_top, old_top_size);
2081 chunk_at_offset(old_top, old_top_size )->size =
wdenk8bde7f72003-06-27 21:31:46 +00002082 SIZE_SZ|PREV_INUSE;
wdenk217c9da2002-10-25 20:35:49 +00002083 chunk_at_offset(old_top, old_top_size + SIZE_SZ)->size =
wdenk8bde7f72003-06-27 21:31:46 +00002084 SIZE_SZ|PREV_INUSE;
wdenk217c9da2002-10-25 20:35:49 +00002085 /* If possible, release the rest. */
2086 if (old_top_size >= MINSIZE)
wdenk8bde7f72003-06-27 21:31:46 +00002087 fREe(chunk2mem(old_top));
wdenk217c9da2002-10-25 20:35:49 +00002088 }
2089 }
2090
2091 if ((unsigned long)sbrked_mem > (unsigned long)max_sbrked_mem)
2092 max_sbrked_mem = sbrked_mem;
2093 if ((unsigned long)(mmapped_mem + sbrked_mem) > (unsigned long)max_total_mem)
2094 max_total_mem = mmapped_mem + sbrked_mem;
2095
2096 /* We always land on a page boundary */
2097 assert(((unsigned long)((char*)top + top_size) & (pagesz - 1)) == 0);
2098}
2099
2100
2101
2102
2103/* Main public routines */
2104
2105
2106/*
2107 Malloc Algorthim:
2108
2109 The requested size is first converted into a usable form, `nb'.
2110 This currently means to add 4 bytes overhead plus possibly more to
2111 obtain 8-byte alignment and/or to obtain a size of at least
2112 MINSIZE (currently 16 bytes), the smallest allocatable size.
2113 (All fits are considered `exact' if they are within MINSIZE bytes.)
2114
2115 From there, the first successful of the following steps is taken:
2116
2117 1. The bin corresponding to the request size is scanned, and if
wdenk8bde7f72003-06-27 21:31:46 +00002118 a chunk of exactly the right size is found, it is taken.
wdenk217c9da2002-10-25 20:35:49 +00002119
2120 2. The most recently remaindered chunk is used if it is big
wdenk8bde7f72003-06-27 21:31:46 +00002121 enough. This is a form of (roving) first fit, used only in
2122 the absence of exact fits. Runs of consecutive requests use
2123 the remainder of the chunk used for the previous such request
2124 whenever possible. This limited use of a first-fit style
2125 allocation strategy tends to give contiguous chunks
2126 coextensive lifetimes, which improves locality and can reduce
2127 fragmentation in the long run.
wdenk217c9da2002-10-25 20:35:49 +00002128
2129 3. Other bins are scanned in increasing size order, using a
wdenk8bde7f72003-06-27 21:31:46 +00002130 chunk big enough to fulfill the request, and splitting off
2131 any remainder. This search is strictly by best-fit; i.e.,
2132 the smallest (with ties going to approximately the least
2133 recently used) chunk that fits is selected.
wdenk217c9da2002-10-25 20:35:49 +00002134
2135 4. If large enough, the chunk bordering the end of memory
wdenk8bde7f72003-06-27 21:31:46 +00002136 (`top') is split off. (This use of `top' is in accord with
2137 the best-fit search rule. In effect, `top' is treated as
2138 larger (and thus less well fitting) than any other available
2139 chunk since it can be extended to be as large as necessary
2140 (up to system limitations).
wdenk217c9da2002-10-25 20:35:49 +00002141
2142 5. If the request size meets the mmap threshold and the
wdenk8bde7f72003-06-27 21:31:46 +00002143 system supports mmap, and there are few enough currently
2144 allocated mmapped regions, and a call to mmap succeeds,
2145 the request is allocated via direct memory mapping.
wdenk217c9da2002-10-25 20:35:49 +00002146
2147 6. Otherwise, the top of memory is extended by
wdenk8bde7f72003-06-27 21:31:46 +00002148 obtaining more space from the system (normally using sbrk,
2149 but definable to anything else via the MORECORE macro).
2150 Memory is gathered from the system (in system page-sized
2151 units) in a way that allows chunks obtained across different
2152 sbrk calls to be consolidated, but does not require
2153 contiguous memory. Thus, it should be safe to intersperse
2154 mallocs with other sbrk calls.
wdenk217c9da2002-10-25 20:35:49 +00002155
2156
2157 All allocations are made from the the `lowest' part of any found
2158 chunk. (The implementation invariant is that prev_inuse is
2159 always true of any allocated chunk; i.e., that each allocated
2160 chunk borders either a previously allocated and still in-use chunk,
2161 or the base of its memory arena.)
2162
2163*/
2164
2165#if __STD_C
2166Void_t* mALLOc(size_t bytes)
2167#else
2168Void_t* mALLOc(bytes) size_t bytes;
2169#endif
2170{
2171 mchunkptr victim; /* inspected/selected chunk */
2172 INTERNAL_SIZE_T victim_size; /* its size */
2173 int idx; /* index for bin traversal */
2174 mbinptr bin; /* associated bin */
2175 mchunkptr remainder; /* remainder from a split */
2176 long remainder_size; /* its size */
2177 int remainder_index; /* its bin index */
2178 unsigned long block; /* block traverser bit */
2179 int startidx; /* first bin of a traversed block */
2180 mchunkptr fwd; /* misc temp for linking */
2181 mchunkptr bck; /* misc temp for linking */
2182 mbinptr q; /* misc temp */
2183
2184 INTERNAL_SIZE_T nb;
2185
Wolfgang Denk27405442010-01-15 11:20:10 +01002186 /* check if mem_malloc_init() was run */
2187 if ((mem_malloc_start == 0) && (mem_malloc_end == 0)) {
2188 /* not initialized yet */
2189 return 0;
2190 }
2191
wdenk217c9da2002-10-25 20:35:49 +00002192 if ((long)bytes < 0) return 0;
2193
2194 nb = request2size(bytes); /* padded request size; */
2195
2196 /* Check for exact match in a bin */
2197
2198 if (is_small_request(nb)) /* Faster version for small requests */
2199 {
2200 idx = smallbin_index(nb);
2201
2202 /* No traversal or size check necessary for small bins. */
2203
2204 q = bin_at(idx);
2205 victim = last(q);
2206
2207 /* Also scan the next one, since it would have a remainder < MINSIZE */
2208 if (victim == q)
2209 {
2210 q = next_bin(q);
2211 victim = last(q);
2212 }
2213 if (victim != q)
2214 {
2215 victim_size = chunksize(victim);
2216 unlink(victim, bck, fwd);
2217 set_inuse_bit_at_offset(victim, victim_size);
2218 check_malloced_chunk(victim, nb);
2219 return chunk2mem(victim);
2220 }
2221
2222 idx += 2; /* Set for bin scan below. We've already scanned 2 bins. */
2223
2224 }
2225 else
2226 {
2227 idx = bin_index(nb);
2228 bin = bin_at(idx);
2229
2230 for (victim = last(bin); victim != bin; victim = victim->bk)
2231 {
2232 victim_size = chunksize(victim);
2233 remainder_size = victim_size - nb;
2234
2235 if (remainder_size >= (long)MINSIZE) /* too big */
2236 {
wdenk8bde7f72003-06-27 21:31:46 +00002237 --idx; /* adjust to rescan below after checking last remainder */
2238 break;
wdenk217c9da2002-10-25 20:35:49 +00002239 }
2240
2241 else if (remainder_size >= 0) /* exact fit */
2242 {
wdenk8bde7f72003-06-27 21:31:46 +00002243 unlink(victim, bck, fwd);
2244 set_inuse_bit_at_offset(victim, victim_size);
2245 check_malloced_chunk(victim, nb);
2246 return chunk2mem(victim);
wdenk217c9da2002-10-25 20:35:49 +00002247 }
2248 }
2249
2250 ++idx;
2251
2252 }
2253
2254 /* Try to use the last split-off remainder */
2255
2256 if ( (victim = last_remainder->fd) != last_remainder)
2257 {
2258 victim_size = chunksize(victim);
2259 remainder_size = victim_size - nb;
2260
2261 if (remainder_size >= (long)MINSIZE) /* re-split */
2262 {
2263 remainder = chunk_at_offset(victim, nb);
2264 set_head(victim, nb | PREV_INUSE);
2265 link_last_remainder(remainder);
2266 set_head(remainder, remainder_size | PREV_INUSE);
2267 set_foot(remainder, remainder_size);
2268 check_malloced_chunk(victim, nb);
2269 return chunk2mem(victim);
2270 }
2271
2272 clear_last_remainder;
2273
2274 if (remainder_size >= 0) /* exhaust */
2275 {
2276 set_inuse_bit_at_offset(victim, victim_size);
2277 check_malloced_chunk(victim, nb);
2278 return chunk2mem(victim);
2279 }
2280
2281 /* Else place in bin */
2282
2283 frontlink(victim, victim_size, remainder_index, bck, fwd);
2284 }
2285
2286 /*
2287 If there are any possibly nonempty big-enough blocks,
2288 search for best fitting chunk by scanning bins in blockwidth units.
2289 */
2290
Stefan Roesef2302d42008-08-06 14:05:38 +02002291 if ( (block = idx2binblock(idx)) <= binblocks_r)
wdenk217c9da2002-10-25 20:35:49 +00002292 {
2293
2294 /* Get to the first marked block */
2295
Stefan Roesef2302d42008-08-06 14:05:38 +02002296 if ( (block & binblocks_r) == 0)
wdenk217c9da2002-10-25 20:35:49 +00002297 {
2298 /* force to an even block boundary */
2299 idx = (idx & ~(BINBLOCKWIDTH - 1)) + BINBLOCKWIDTH;
2300 block <<= 1;
Stefan Roesef2302d42008-08-06 14:05:38 +02002301 while ((block & binblocks_r) == 0)
wdenk217c9da2002-10-25 20:35:49 +00002302 {
wdenk8bde7f72003-06-27 21:31:46 +00002303 idx += BINBLOCKWIDTH;
2304 block <<= 1;
wdenk217c9da2002-10-25 20:35:49 +00002305 }
2306 }
2307
2308 /* For each possibly nonempty block ... */
2309 for (;;)
2310 {
2311 startidx = idx; /* (track incomplete blocks) */
2312 q = bin = bin_at(idx);
2313
2314 /* For each bin in this block ... */
2315 do
2316 {
wdenk8bde7f72003-06-27 21:31:46 +00002317 /* Find and use first big enough chunk ... */
wdenk217c9da2002-10-25 20:35:49 +00002318
wdenk8bde7f72003-06-27 21:31:46 +00002319 for (victim = last(bin); victim != bin; victim = victim->bk)
2320 {
2321 victim_size = chunksize(victim);
2322 remainder_size = victim_size - nb;
wdenk217c9da2002-10-25 20:35:49 +00002323
wdenk8bde7f72003-06-27 21:31:46 +00002324 if (remainder_size >= (long)MINSIZE) /* split */
2325 {
2326 remainder = chunk_at_offset(victim, nb);
2327 set_head(victim, nb | PREV_INUSE);
2328 unlink(victim, bck, fwd);
2329 link_last_remainder(remainder);
2330 set_head(remainder, remainder_size | PREV_INUSE);
2331 set_foot(remainder, remainder_size);
2332 check_malloced_chunk(victim, nb);
2333 return chunk2mem(victim);
2334 }
wdenk217c9da2002-10-25 20:35:49 +00002335
wdenk8bde7f72003-06-27 21:31:46 +00002336 else if (remainder_size >= 0) /* take */
2337 {
2338 set_inuse_bit_at_offset(victim, victim_size);
2339 unlink(victim, bck, fwd);
2340 check_malloced_chunk(victim, nb);
2341 return chunk2mem(victim);
2342 }
wdenk217c9da2002-10-25 20:35:49 +00002343
wdenk8bde7f72003-06-27 21:31:46 +00002344 }
wdenk217c9da2002-10-25 20:35:49 +00002345
2346 bin = next_bin(bin);
2347
2348 } while ((++idx & (BINBLOCKWIDTH - 1)) != 0);
2349
2350 /* Clear out the block bit. */
2351
2352 do /* Possibly backtrack to try to clear a partial block */
2353 {
wdenk8bde7f72003-06-27 21:31:46 +00002354 if ((startidx & (BINBLOCKWIDTH - 1)) == 0)
2355 {
Stefan Roesef2302d42008-08-06 14:05:38 +02002356 av_[1] = (mbinptr)(binblocks_r & ~block);
wdenk8bde7f72003-06-27 21:31:46 +00002357 break;
2358 }
2359 --startidx;
wdenk217c9da2002-10-25 20:35:49 +00002360 q = prev_bin(q);
2361 } while (first(q) == q);
2362
2363 /* Get to the next possibly nonempty block */
2364
Stefan Roesef2302d42008-08-06 14:05:38 +02002365 if ( (block <<= 1) <= binblocks_r && (block != 0) )
wdenk217c9da2002-10-25 20:35:49 +00002366 {
Stefan Roesef2302d42008-08-06 14:05:38 +02002367 while ((block & binblocks_r) == 0)
wdenk8bde7f72003-06-27 21:31:46 +00002368 {
2369 idx += BINBLOCKWIDTH;
2370 block <<= 1;
2371 }
wdenk217c9da2002-10-25 20:35:49 +00002372 }
2373 else
wdenk8bde7f72003-06-27 21:31:46 +00002374 break;
wdenk217c9da2002-10-25 20:35:49 +00002375 }
2376 }
2377
2378
2379 /* Try to use top chunk */
2380
2381 /* Require that there be a remainder, ensuring top always exists */
2382 if ( (remainder_size = chunksize(top) - nb) < (long)MINSIZE)
2383 {
2384
2385#if HAVE_MMAP
2386 /* If big and would otherwise need to extend, try to use mmap instead */
2387 if ((unsigned long)nb >= (unsigned long)mmap_threshold &&
wdenk8bde7f72003-06-27 21:31:46 +00002388 (victim = mmap_chunk(nb)) != 0)
wdenk217c9da2002-10-25 20:35:49 +00002389 return chunk2mem(victim);
2390#endif
2391
2392 /* Try to extend */
2393 malloc_extend_top(nb);
2394 if ( (remainder_size = chunksize(top) - nb) < (long)MINSIZE)
2395 return 0; /* propagate failure */
2396 }
2397
2398 victim = top;
2399 set_head(victim, nb | PREV_INUSE);
2400 top = chunk_at_offset(victim, nb);
2401 set_head(top, remainder_size | PREV_INUSE);
2402 check_malloced_chunk(victim, nb);
2403 return chunk2mem(victim);
2404
2405}
2406
2407
2408
2409
2410/*
2411
2412 free() algorithm :
2413
2414 cases:
2415
2416 1. free(0) has no effect.
2417
2418 2. If the chunk was allocated via mmap, it is release via munmap().
2419
2420 3. If a returned chunk borders the current high end of memory,
wdenk8bde7f72003-06-27 21:31:46 +00002421 it is consolidated into the top, and if the total unused
2422 topmost memory exceeds the trim threshold, malloc_trim is
2423 called.
wdenk217c9da2002-10-25 20:35:49 +00002424
2425 4. Other chunks are consolidated as they arrive, and
wdenk8bde7f72003-06-27 21:31:46 +00002426 placed in corresponding bins. (This includes the case of
2427 consolidating with the current `last_remainder').
wdenk217c9da2002-10-25 20:35:49 +00002428
2429*/
2430
2431
2432#if __STD_C
2433void fREe(Void_t* mem)
2434#else
2435void fREe(mem) Void_t* mem;
2436#endif
2437{
2438 mchunkptr p; /* chunk corresponding to mem */
2439 INTERNAL_SIZE_T hd; /* its head field */
2440 INTERNAL_SIZE_T sz; /* its size */
2441 int idx; /* its bin index */
2442 mchunkptr next; /* next contiguous chunk */
2443 INTERNAL_SIZE_T nextsz; /* its size */
2444 INTERNAL_SIZE_T prevsz; /* size of previous contiguous chunk */
2445 mchunkptr bck; /* misc temp for linking */
2446 mchunkptr fwd; /* misc temp for linking */
2447 int islr; /* track whether merging with last_remainder */
2448
2449 if (mem == 0) /* free(0) has no effect */
2450 return;
2451
2452 p = mem2chunk(mem);
2453 hd = p->size;
2454
2455#if HAVE_MMAP
2456 if (hd & IS_MMAPPED) /* release mmapped memory. */
2457 {
2458 munmap_chunk(p);
2459 return;
2460 }
2461#endif
2462
2463 check_inuse_chunk(p);
2464
2465 sz = hd & ~PREV_INUSE;
2466 next = chunk_at_offset(p, sz);
2467 nextsz = chunksize(next);
2468
2469 if (next == top) /* merge with top */
2470 {
2471 sz += nextsz;
2472
2473 if (!(hd & PREV_INUSE)) /* consolidate backward */
2474 {
2475 prevsz = p->prev_size;
2476 p = chunk_at_offset(p, -((long) prevsz));
2477 sz += prevsz;
2478 unlink(p, bck, fwd);
2479 }
2480
2481 set_head(p, sz | PREV_INUSE);
2482 top = p;
2483 if ((unsigned long)(sz) >= (unsigned long)trim_threshold)
2484 malloc_trim(top_pad);
2485 return;
2486 }
2487
2488 set_head(next, nextsz); /* clear inuse bit */
2489
2490 islr = 0;
2491
2492 if (!(hd & PREV_INUSE)) /* consolidate backward */
2493 {
2494 prevsz = p->prev_size;
2495 p = chunk_at_offset(p, -((long) prevsz));
2496 sz += prevsz;
2497
2498 if (p->fd == last_remainder) /* keep as last_remainder */
2499 islr = 1;
2500 else
2501 unlink(p, bck, fwd);
2502 }
2503
2504 if (!(inuse_bit_at_offset(next, nextsz))) /* consolidate forward */
2505 {
2506 sz += nextsz;
2507
2508 if (!islr && next->fd == last_remainder) /* re-insert last_remainder */
2509 {
2510 islr = 1;
2511 link_last_remainder(p);
2512 }
2513 else
2514 unlink(next, bck, fwd);
2515 }
2516
2517
2518 set_head(p, sz | PREV_INUSE);
2519 set_foot(p, sz);
2520 if (!islr)
2521 frontlink(p, sz, idx, bck, fwd);
2522}
2523
2524
2525
2526
2527
2528/*
2529
2530 Realloc algorithm:
2531
2532 Chunks that were obtained via mmap cannot be extended or shrunk
2533 unless HAVE_MREMAP is defined, in which case mremap is used.
2534 Otherwise, if their reallocation is for additional space, they are
2535 copied. If for less, they are just left alone.
2536
2537 Otherwise, if the reallocation is for additional space, and the
2538 chunk can be extended, it is, else a malloc-copy-free sequence is
2539 taken. There are several different ways that a chunk could be
2540 extended. All are tried:
2541
2542 * Extending forward into following adjacent free chunk.
2543 * Shifting backwards, joining preceding adjacent space
2544 * Both shifting backwards and extending forward.
2545 * Extending into newly sbrked space
2546
2547 Unless the #define REALLOC_ZERO_BYTES_FREES is set, realloc with a
2548 size argument of zero (re)allocates a minimum-sized chunk.
2549
2550 If the reallocation is for less space, and the new request is for
2551 a `small' (<512 bytes) size, then the newly unused space is lopped
2552 off and freed.
2553
2554 The old unix realloc convention of allowing the last-free'd chunk
2555 to be used as an argument to realloc is no longer supported.
2556 I don't know of any programs still relying on this feature,
2557 and allowing it would also allow too many other incorrect
2558 usages of realloc to be sensible.
2559
2560
2561*/
2562
2563
2564#if __STD_C
2565Void_t* rEALLOc(Void_t* oldmem, size_t bytes)
2566#else
2567Void_t* rEALLOc(oldmem, bytes) Void_t* oldmem; size_t bytes;
2568#endif
2569{
2570 INTERNAL_SIZE_T nb; /* padded request size */
2571
2572 mchunkptr oldp; /* chunk corresponding to oldmem */
2573 INTERNAL_SIZE_T oldsize; /* its size */
2574
2575 mchunkptr newp; /* chunk to return */
2576 INTERNAL_SIZE_T newsize; /* its size */
2577 Void_t* newmem; /* corresponding user mem */
2578
2579 mchunkptr next; /* next contiguous chunk after oldp */
2580 INTERNAL_SIZE_T nextsize; /* its size */
2581
2582 mchunkptr prev; /* previous contiguous chunk before oldp */
2583 INTERNAL_SIZE_T prevsize; /* its size */
2584
2585 mchunkptr remainder; /* holds split off extra space from newp */
2586 INTERNAL_SIZE_T remainder_size; /* its size */
2587
2588 mchunkptr bck; /* misc temp for linking */
2589 mchunkptr fwd; /* misc temp for linking */
2590
2591#ifdef REALLOC_ZERO_BYTES_FREES
2592 if (bytes == 0) { fREe(oldmem); return 0; }
2593#endif
2594
2595 if ((long)bytes < 0) return 0;
2596
2597 /* realloc of null is supposed to be same as malloc */
2598 if (oldmem == 0) return mALLOc(bytes);
2599
2600 newp = oldp = mem2chunk(oldmem);
2601 newsize = oldsize = chunksize(oldp);
2602
2603
2604 nb = request2size(bytes);
2605
2606#if HAVE_MMAP
2607 if (chunk_is_mmapped(oldp))
2608 {
2609#if HAVE_MREMAP
2610 newp = mremap_chunk(oldp, nb);
2611 if(newp) return chunk2mem(newp);
2612#endif
2613 /* Note the extra SIZE_SZ overhead. */
2614 if(oldsize - SIZE_SZ >= nb) return oldmem; /* do nothing */
2615 /* Must alloc, copy, free. */
2616 newmem = mALLOc(bytes);
2617 if (newmem == 0) return 0; /* propagate failure */
2618 MALLOC_COPY(newmem, oldmem, oldsize - 2*SIZE_SZ);
2619 munmap_chunk(oldp);
2620 return newmem;
2621 }
2622#endif
2623
2624 check_inuse_chunk(oldp);
2625
2626 if ((long)(oldsize) < (long)(nb))
2627 {
2628
2629 /* Try expanding forward */
2630
2631 next = chunk_at_offset(oldp, oldsize);
2632 if (next == top || !inuse(next))
2633 {
2634 nextsize = chunksize(next);
2635
2636 /* Forward into top only if a remainder */
2637 if (next == top)
2638 {
wdenk8bde7f72003-06-27 21:31:46 +00002639 if ((long)(nextsize + newsize) >= (long)(nb + MINSIZE))
2640 {
2641 newsize += nextsize;
2642 top = chunk_at_offset(oldp, nb);
2643 set_head(top, (newsize - nb) | PREV_INUSE);
2644 set_head_size(oldp, nb);
2645 return chunk2mem(oldp);
2646 }
wdenk217c9da2002-10-25 20:35:49 +00002647 }
2648
2649 /* Forward into next chunk */
2650 else if (((long)(nextsize + newsize) >= (long)(nb)))
2651 {
wdenk8bde7f72003-06-27 21:31:46 +00002652 unlink(next, bck, fwd);
2653 newsize += nextsize;
2654 goto split;
wdenk217c9da2002-10-25 20:35:49 +00002655 }
2656 }
2657 else
2658 {
2659 next = 0;
2660 nextsize = 0;
2661 }
2662
2663 /* Try shifting backwards. */
2664
2665 if (!prev_inuse(oldp))
2666 {
2667 prev = prev_chunk(oldp);
2668 prevsize = chunksize(prev);
2669
2670 /* try forward + backward first to save a later consolidation */
2671
2672 if (next != 0)
2673 {
wdenk8bde7f72003-06-27 21:31:46 +00002674 /* into top */
2675 if (next == top)
2676 {
2677 if ((long)(nextsize + prevsize + newsize) >= (long)(nb + MINSIZE))
2678 {
2679 unlink(prev, bck, fwd);
2680 newp = prev;
2681 newsize += prevsize + nextsize;
2682 newmem = chunk2mem(newp);
2683 MALLOC_COPY(newmem, oldmem, oldsize - SIZE_SZ);
2684 top = chunk_at_offset(newp, nb);
2685 set_head(top, (newsize - nb) | PREV_INUSE);
2686 set_head_size(newp, nb);
2687 return newmem;
2688 }
2689 }
wdenk217c9da2002-10-25 20:35:49 +00002690
wdenk8bde7f72003-06-27 21:31:46 +00002691 /* into next chunk */
2692 else if (((long)(nextsize + prevsize + newsize) >= (long)(nb)))
2693 {
2694 unlink(next, bck, fwd);
2695 unlink(prev, bck, fwd);
2696 newp = prev;
2697 newsize += nextsize + prevsize;
2698 newmem = chunk2mem(newp);
2699 MALLOC_COPY(newmem, oldmem, oldsize - SIZE_SZ);
2700 goto split;
2701 }
wdenk217c9da2002-10-25 20:35:49 +00002702 }
2703
2704 /* backward only */
2705 if (prev != 0 && (long)(prevsize + newsize) >= (long)nb)
2706 {
wdenk8bde7f72003-06-27 21:31:46 +00002707 unlink(prev, bck, fwd);
2708 newp = prev;
2709 newsize += prevsize;
2710 newmem = chunk2mem(newp);
2711 MALLOC_COPY(newmem, oldmem, oldsize - SIZE_SZ);
2712 goto split;
wdenk217c9da2002-10-25 20:35:49 +00002713 }
2714 }
2715
2716 /* Must allocate */
2717
2718 newmem = mALLOc (bytes);
2719
2720 if (newmem == 0) /* propagate failure */
2721 return 0;
2722
2723 /* Avoid copy if newp is next chunk after oldp. */
2724 /* (This can only happen when new chunk is sbrk'ed.) */
2725
2726 if ( (newp = mem2chunk(newmem)) == next_chunk(oldp))
2727 {
2728 newsize += chunksize(newp);
2729 newp = oldp;
2730 goto split;
2731 }
2732
2733 /* Otherwise copy, free, and exit */
2734 MALLOC_COPY(newmem, oldmem, oldsize - SIZE_SZ);
2735 fREe(oldmem);
2736 return newmem;
2737 }
2738
2739
2740 split: /* split off extra room in old or expanded chunk */
2741
2742 if (newsize - nb >= MINSIZE) /* split off remainder */
2743 {
2744 remainder = chunk_at_offset(newp, nb);
2745 remainder_size = newsize - nb;
2746 set_head_size(newp, nb);
2747 set_head(remainder, remainder_size | PREV_INUSE);
2748 set_inuse_bit_at_offset(remainder, remainder_size);
2749 fREe(chunk2mem(remainder)); /* let free() deal with it */
2750 }
2751 else
2752 {
2753 set_head_size(newp, newsize);
2754 set_inuse_bit_at_offset(newp, newsize);
2755 }
2756
2757 check_inuse_chunk(newp);
2758 return chunk2mem(newp);
2759}
2760
2761
2762
2763
2764/*
2765
2766 memalign algorithm:
2767
2768 memalign requests more than enough space from malloc, finds a spot
2769 within that chunk that meets the alignment request, and then
2770 possibly frees the leading and trailing space.
2771
2772 The alignment argument must be a power of two. This property is not
2773 checked by memalign, so misuse may result in random runtime errors.
2774
2775 8-byte alignment is guaranteed by normal malloc calls, so don't
2776 bother calling memalign with an argument of 8 or less.
2777
2778 Overreliance on memalign is a sure way to fragment space.
2779
2780*/
2781
2782
2783#if __STD_C
2784Void_t* mEMALIGn(size_t alignment, size_t bytes)
2785#else
2786Void_t* mEMALIGn(alignment, bytes) size_t alignment; size_t bytes;
2787#endif
2788{
2789 INTERNAL_SIZE_T nb; /* padded request size */
2790 char* m; /* memory returned by malloc call */
2791 mchunkptr p; /* corresponding chunk */
2792 char* brk; /* alignment point within p */
2793 mchunkptr newp; /* chunk to return */
2794 INTERNAL_SIZE_T newsize; /* its size */
2795 INTERNAL_SIZE_T leadsize; /* leading space befor alignment point */
2796 mchunkptr remainder; /* spare room at end to split off */
2797 long remainder_size; /* its size */
2798
2799 if ((long)bytes < 0) return 0;
2800
2801 /* If need less alignment than we give anyway, just relay to malloc */
2802
2803 if (alignment <= MALLOC_ALIGNMENT) return mALLOc(bytes);
2804
2805 /* Otherwise, ensure that it is at least a minimum chunk size */
2806
2807 if (alignment < MINSIZE) alignment = MINSIZE;
2808
2809 /* Call malloc with worst case padding to hit alignment. */
2810
2811 nb = request2size(bytes);
2812 m = (char*)(mALLOc(nb + alignment + MINSIZE));
2813
2814 if (m == 0) return 0; /* propagate failure */
2815
2816 p = mem2chunk(m);
2817
2818 if ((((unsigned long)(m)) % alignment) == 0) /* aligned */
2819 {
2820#if HAVE_MMAP
2821 if(chunk_is_mmapped(p))
2822 return chunk2mem(p); /* nothing more to do */
2823#endif
2824 }
2825 else /* misaligned */
2826 {
2827 /*
2828 Find an aligned spot inside chunk.
2829 Since we need to give back leading space in a chunk of at
2830 least MINSIZE, if the first calculation places us at
2831 a spot with less than MINSIZE leader, we can move to the
2832 next aligned spot -- we've allocated enough total room so that
2833 this is always possible.
2834 */
2835
2836 brk = (char*)mem2chunk(((unsigned long)(m + alignment - 1)) & -((signed) alignment));
2837 if ((long)(brk - (char*)(p)) < MINSIZE) brk = brk + alignment;
2838
2839 newp = (mchunkptr)brk;
2840 leadsize = brk - (char*)(p);
2841 newsize = chunksize(p) - leadsize;
2842
2843#if HAVE_MMAP
2844 if(chunk_is_mmapped(p))
2845 {
2846 newp->prev_size = p->prev_size + leadsize;
2847 set_head(newp, newsize|IS_MMAPPED);
2848 return chunk2mem(newp);
2849 }
2850#endif
2851
2852 /* give back leader, use the rest */
2853
2854 set_head(newp, newsize | PREV_INUSE);
2855 set_inuse_bit_at_offset(newp, newsize);
2856 set_head_size(p, leadsize);
2857 fREe(chunk2mem(p));
2858 p = newp;
2859
2860 assert (newsize >= nb && (((unsigned long)(chunk2mem(p))) % alignment) == 0);
2861 }
2862
2863 /* Also give back spare room at the end */
2864
2865 remainder_size = chunksize(p) - nb;
2866
2867 if (remainder_size >= (long)MINSIZE)
2868 {
2869 remainder = chunk_at_offset(p, nb);
2870 set_head(remainder, remainder_size | PREV_INUSE);
2871 set_head_size(p, nb);
2872 fREe(chunk2mem(remainder));
2873 }
2874
2875 check_inuse_chunk(p);
2876 return chunk2mem(p);
2877
2878}
2879
2880
2881
2882
2883/*
2884 valloc just invokes memalign with alignment argument equal
2885 to the page size of the system (or as near to this as can
2886 be figured out from all the includes/defines above.)
2887*/
2888
2889#if __STD_C
2890Void_t* vALLOc(size_t bytes)
2891#else
2892Void_t* vALLOc(bytes) size_t bytes;
2893#endif
2894{
2895 return mEMALIGn (malloc_getpagesize, bytes);
2896}
2897
2898/*
2899 pvalloc just invokes valloc for the nearest pagesize
2900 that will accommodate request
2901*/
2902
2903
2904#if __STD_C
2905Void_t* pvALLOc(size_t bytes)
2906#else
2907Void_t* pvALLOc(bytes) size_t bytes;
2908#endif
2909{
2910 size_t pagesize = malloc_getpagesize;
2911 return mEMALIGn (pagesize, (bytes + pagesize - 1) & ~(pagesize - 1));
2912}
2913
2914/*
2915
2916 calloc calls malloc, then zeroes out the allocated chunk.
2917
2918*/
2919
2920#if __STD_C
2921Void_t* cALLOc(size_t n, size_t elem_size)
2922#else
2923Void_t* cALLOc(n, elem_size) size_t n; size_t elem_size;
2924#endif
2925{
2926 mchunkptr p;
2927 INTERNAL_SIZE_T csz;
2928
2929 INTERNAL_SIZE_T sz = n * elem_size;
2930
2931
2932 /* check if expand_top called, in which case don't need to clear */
2933#if MORECORE_CLEARS
2934 mchunkptr oldtop = top;
2935 INTERNAL_SIZE_T oldtopsize = chunksize(top);
2936#endif
2937 Void_t* mem = mALLOc (sz);
2938
2939 if ((long)n < 0) return 0;
2940
2941 if (mem == 0)
2942 return 0;
2943 else
2944 {
2945 p = mem2chunk(mem);
2946
2947 /* Two optional cases in which clearing not necessary */
2948
2949
2950#if HAVE_MMAP
2951 if (chunk_is_mmapped(p)) return mem;
2952#endif
2953
2954 csz = chunksize(p);
2955
2956#if MORECORE_CLEARS
2957 if (p == oldtop && csz > oldtopsize)
2958 {
2959 /* clear only the bytes from non-freshly-sbrked memory */
2960 csz = oldtopsize;
2961 }
2962#endif
2963
2964 MALLOC_ZERO(mem, csz - SIZE_SZ);
2965 return mem;
2966 }
2967}
2968
2969/*
2970
2971 cfree just calls free. It is needed/defined on some systems
2972 that pair it with calloc, presumably for odd historical reasons.
2973
2974*/
2975
2976#if !defined(INTERNAL_LINUX_C_LIB) || !defined(__ELF__)
2977#if __STD_C
2978void cfree(Void_t *mem)
2979#else
2980void cfree(mem) Void_t *mem;
2981#endif
2982{
2983 fREe(mem);
2984}
2985#endif
2986
2987
2988
2989/*
2990
2991 Malloc_trim gives memory back to the system (via negative
2992 arguments to sbrk) if there is unused memory at the `high' end of
2993 the malloc pool. You can call this after freeing large blocks of
2994 memory to potentially reduce the system-level memory requirements
2995 of a program. However, it cannot guarantee to reduce memory. Under
2996 some allocation patterns, some large free blocks of memory will be
2997 locked between two used chunks, so they cannot be given back to
2998 the system.
2999
3000 The `pad' argument to malloc_trim represents the amount of free
3001 trailing space to leave untrimmed. If this argument is zero,
3002 only the minimum amount of memory to maintain internal data
3003 structures will be left (one page or less). Non-zero arguments
3004 can be supplied to maintain enough trailing space to service
3005 future expected allocations without having to re-obtain memory
3006 from the system.
3007
3008 Malloc_trim returns 1 if it actually released any memory, else 0.
3009
3010*/
3011
3012#if __STD_C
3013int malloc_trim(size_t pad)
3014#else
3015int malloc_trim(pad) size_t pad;
3016#endif
3017{
3018 long top_size; /* Amount of top-most memory */
3019 long extra; /* Amount to release */
3020 char* current_brk; /* address returned by pre-check sbrk call */
3021 char* new_brk; /* address returned by negative sbrk call */
3022
3023 unsigned long pagesz = malloc_getpagesize;
3024
3025 top_size = chunksize(top);
3026 extra = ((top_size - pad - MINSIZE + (pagesz-1)) / pagesz - 1) * pagesz;
3027
3028 if (extra < (long)pagesz) /* Not enough memory to release */
3029 return 0;
3030
3031 else
3032 {
3033 /* Test to make sure no one else called sbrk */
3034 current_brk = (char*)(MORECORE (0));
3035 if (current_brk != (char*)(top) + top_size)
3036 return 0; /* Apparently we don't own memory; must fail */
3037
3038 else
3039 {
3040 new_brk = (char*)(MORECORE (-extra));
3041
3042 if (new_brk == (char*)(MORECORE_FAILURE)) /* sbrk failed? */
3043 {
wdenk8bde7f72003-06-27 21:31:46 +00003044 /* Try to figure out what we have */
3045 current_brk = (char*)(MORECORE (0));
3046 top_size = current_brk - (char*)top;
3047 if (top_size >= (long)MINSIZE) /* if not, we are very very dead! */
3048 {
3049 sbrked_mem = current_brk - sbrk_base;
3050 set_head(top, top_size | PREV_INUSE);
3051 }
3052 check_chunk(top);
3053 return 0;
wdenk217c9da2002-10-25 20:35:49 +00003054 }
3055
3056 else
3057 {
wdenk8bde7f72003-06-27 21:31:46 +00003058 /* Success. Adjust top accordingly. */
3059 set_head(top, (top_size - extra) | PREV_INUSE);
3060 sbrked_mem -= extra;
3061 check_chunk(top);
3062 return 1;
wdenk217c9da2002-10-25 20:35:49 +00003063 }
3064 }
3065 }
3066}
3067
3068
3069
3070/*
3071 malloc_usable_size:
3072
3073 This routine tells you how many bytes you can actually use in an
3074 allocated chunk, which may be more than you requested (although
3075 often not). You can use this many bytes without worrying about
3076 overwriting other allocated objects. Not a particularly great
3077 programming practice, but still sometimes useful.
3078
3079*/
3080
3081#if __STD_C
3082size_t malloc_usable_size(Void_t* mem)
3083#else
3084size_t malloc_usable_size(mem) Void_t* mem;
3085#endif
3086{
3087 mchunkptr p;
3088 if (mem == 0)
3089 return 0;
3090 else
3091 {
3092 p = mem2chunk(mem);
3093 if(!chunk_is_mmapped(p))
3094 {
3095 if (!inuse(p)) return 0;
3096 check_inuse_chunk(p);
3097 return chunksize(p) - SIZE_SZ;
3098 }
3099 return chunksize(p) - 2*SIZE_SZ;
3100 }
3101}
3102
3103
3104
3105
3106/* Utility to update current_mallinfo for malloc_stats and mallinfo() */
3107
Wolfgang Denkea882ba2010-06-20 23:33:59 +02003108#ifdef DEBUG
wdenk217c9da2002-10-25 20:35:49 +00003109static void malloc_update_mallinfo()
3110{
3111 int i;
3112 mbinptr b;
3113 mchunkptr p;
3114#ifdef DEBUG
3115 mchunkptr q;
3116#endif
3117
3118 INTERNAL_SIZE_T avail = chunksize(top);
3119 int navail = ((long)(avail) >= (long)MINSIZE)? 1 : 0;
3120
3121 for (i = 1; i < NAV; ++i)
3122 {
3123 b = bin_at(i);
3124 for (p = last(b); p != b; p = p->bk)
3125 {
3126#ifdef DEBUG
3127 check_free_chunk(p);
3128 for (q = next_chunk(p);
wdenk8bde7f72003-06-27 21:31:46 +00003129 q < top && inuse(q) && (long)(chunksize(q)) >= (long)MINSIZE;
3130 q = next_chunk(q))
3131 check_inuse_chunk(q);
wdenk217c9da2002-10-25 20:35:49 +00003132#endif
3133 avail += chunksize(p);
3134 navail++;
3135 }
3136 }
3137
3138 current_mallinfo.ordblks = navail;
3139 current_mallinfo.uordblks = sbrked_mem - avail;
3140 current_mallinfo.fordblks = avail;
3141 current_mallinfo.hblks = n_mmaps;
3142 current_mallinfo.hblkhd = mmapped_mem;
3143 current_mallinfo.keepcost = chunksize(top);
3144
3145}
Wolfgang Denkea882ba2010-06-20 23:33:59 +02003146#endif /* DEBUG */
wdenk217c9da2002-10-25 20:35:49 +00003147
3148
3149
3150/*
3151
3152 malloc_stats:
3153
3154 Prints on the amount of space obtain from the system (both
3155 via sbrk and mmap), the maximum amount (which may be more than
3156 current if malloc_trim and/or munmap got called), the maximum
3157 number of simultaneous mmap regions used, and the current number
3158 of bytes allocated via malloc (or realloc, etc) but not yet
3159 freed. (Note that this is the number of bytes allocated, not the
3160 number requested. It will be larger than the number requested
3161 because of alignment and bookkeeping overhead.)
3162
3163*/
3164
Wolfgang Denkea882ba2010-06-20 23:33:59 +02003165#ifdef DEBUG
wdenk217c9da2002-10-25 20:35:49 +00003166void malloc_stats()
3167{
3168 malloc_update_mallinfo();
3169 printf("max system bytes = %10u\n",
wdenk8bde7f72003-06-27 21:31:46 +00003170 (unsigned int)(max_total_mem));
wdenk217c9da2002-10-25 20:35:49 +00003171 printf("system bytes = %10u\n",
wdenk8bde7f72003-06-27 21:31:46 +00003172 (unsigned int)(sbrked_mem + mmapped_mem));
wdenk217c9da2002-10-25 20:35:49 +00003173 printf("in use bytes = %10u\n",
wdenk8bde7f72003-06-27 21:31:46 +00003174 (unsigned int)(current_mallinfo.uordblks + mmapped_mem));
wdenk217c9da2002-10-25 20:35:49 +00003175#if HAVE_MMAP
3176 printf("max mmap regions = %10u\n",
wdenk8bde7f72003-06-27 21:31:46 +00003177 (unsigned int)max_n_mmaps);
wdenk217c9da2002-10-25 20:35:49 +00003178#endif
3179}
Wolfgang Denkea882ba2010-06-20 23:33:59 +02003180#endif /* DEBUG */
wdenk217c9da2002-10-25 20:35:49 +00003181
3182/*
3183 mallinfo returns a copy of updated current mallinfo.
3184*/
3185
Wolfgang Denkea882ba2010-06-20 23:33:59 +02003186#ifdef DEBUG
wdenk217c9da2002-10-25 20:35:49 +00003187struct mallinfo mALLINFo()
3188{
3189 malloc_update_mallinfo();
3190 return current_mallinfo;
3191}
Wolfgang Denkea882ba2010-06-20 23:33:59 +02003192#endif /* DEBUG */
wdenk217c9da2002-10-25 20:35:49 +00003193
3194
3195
3196
3197/*
3198 mallopt:
3199
3200 mallopt is the general SVID/XPG interface to tunable parameters.
3201 The format is to provide a (parameter-number, parameter-value) pair.
3202 mallopt then sets the corresponding parameter to the argument
3203 value if it can (i.e., so long as the value is meaningful),
3204 and returns 1 if successful else 0.
3205
3206 See descriptions of tunable parameters above.
3207
3208*/
3209
3210#if __STD_C
3211int mALLOPt(int param_number, int value)
3212#else
3213int mALLOPt(param_number, value) int param_number; int value;
3214#endif
3215{
3216 switch(param_number)
3217 {
3218 case M_TRIM_THRESHOLD:
3219 trim_threshold = value; return 1;
3220 case M_TOP_PAD:
3221 top_pad = value; return 1;
3222 case M_MMAP_THRESHOLD:
3223 mmap_threshold = value; return 1;
3224 case M_MMAP_MAX:
3225#if HAVE_MMAP
3226 n_mmaps_max = value; return 1;
3227#else
3228 if (value != 0) return 0; else n_mmaps_max = value; return 1;
3229#endif
3230
3231 default:
3232 return 0;
3233 }
3234}
3235
3236/*
3237
3238History:
3239
3240 V2.6.6 Sun Dec 5 07:42:19 1999 Doug Lea (dl at gee)
3241 * return null for negative arguments
3242 * Added Several WIN32 cleanups from Martin C. Fong <mcfong@yahoo.com>
wdenk8bde7f72003-06-27 21:31:46 +00003243 * Add 'LACKS_SYS_PARAM_H' for those systems without 'sys/param.h'
3244 (e.g. WIN32 platforms)
3245 * Cleanup up header file inclusion for WIN32 platforms
3246 * Cleanup code to avoid Microsoft Visual C++ compiler complaints
3247 * Add 'USE_DL_PREFIX' to quickly allow co-existence with existing
3248 memory allocation routines
3249 * Set 'malloc_getpagesize' for WIN32 platforms (needs more work)
3250 * Use 'assert' rather than 'ASSERT' in WIN32 code to conform to
wdenk217c9da2002-10-25 20:35:49 +00003251 usage of 'assert' in non-WIN32 code
wdenk8bde7f72003-06-27 21:31:46 +00003252 * Improve WIN32 'sbrk()' emulation's 'findRegion()' routine to
3253 avoid infinite loop
wdenk217c9da2002-10-25 20:35:49 +00003254 * Always call 'fREe()' rather than 'free()'
3255
3256 V2.6.5 Wed Jun 17 15:57:31 1998 Doug Lea (dl at gee)
3257 * Fixed ordering problem with boundary-stamping
3258
3259 V2.6.3 Sun May 19 08:17:58 1996 Doug Lea (dl at gee)
3260 * Added pvalloc, as recommended by H.J. Liu
3261 * Added 64bit pointer support mainly from Wolfram Gloger
3262 * Added anonymously donated WIN32 sbrk emulation
3263 * Malloc, calloc, getpagesize: add optimizations from Raymond Nijssen
3264 * malloc_extend_top: fix mask error that caused wastage after
wdenk8bde7f72003-06-27 21:31:46 +00003265 foreign sbrks
wdenk217c9da2002-10-25 20:35:49 +00003266 * Add linux mremap support code from HJ Liu
3267
3268 V2.6.2 Tue Dec 5 06:52:55 1995 Doug Lea (dl at gee)
3269 * Integrated most documentation with the code.
3270 * Add support for mmap, with help from
wdenk8bde7f72003-06-27 21:31:46 +00003271 Wolfram Gloger (Gloger@lrz.uni-muenchen.de).
wdenk217c9da2002-10-25 20:35:49 +00003272 * Use last_remainder in more cases.
3273 * Pack bins using idea from colin@nyx10.cs.du.edu
3274 * Use ordered bins instead of best-fit threshhold
3275 * Eliminate block-local decls to simplify tracing and debugging.
3276 * Support another case of realloc via move into top
3277 * Fix error occuring when initial sbrk_base not word-aligned.
3278 * Rely on page size for units instead of SBRK_UNIT to
wdenk8bde7f72003-06-27 21:31:46 +00003279 avoid surprises about sbrk alignment conventions.
wdenk217c9da2002-10-25 20:35:49 +00003280 * Add mallinfo, mallopt. Thanks to Raymond Nijssen
wdenk8bde7f72003-06-27 21:31:46 +00003281 (raymond@es.ele.tue.nl) for the suggestion.
wdenk217c9da2002-10-25 20:35:49 +00003282 * Add `pad' argument to malloc_trim and top_pad mallopt parameter.
3283 * More precautions for cases where other routines call sbrk,
wdenk8bde7f72003-06-27 21:31:46 +00003284 courtesy of Wolfram Gloger (Gloger@lrz.uni-muenchen.de).
wdenk217c9da2002-10-25 20:35:49 +00003285 * Added macros etc., allowing use in linux libc from
wdenk8bde7f72003-06-27 21:31:46 +00003286 H.J. Lu (hjl@gnu.ai.mit.edu)
wdenk217c9da2002-10-25 20:35:49 +00003287 * Inverted this history list
3288
3289 V2.6.1 Sat Dec 2 14:10:57 1995 Doug Lea (dl at gee)
3290 * Re-tuned and fixed to behave more nicely with V2.6.0 changes.
3291 * Removed all preallocation code since under current scheme
wdenk8bde7f72003-06-27 21:31:46 +00003292 the work required to undo bad preallocations exceeds
3293 the work saved in good cases for most test programs.
wdenk217c9da2002-10-25 20:35:49 +00003294 * No longer use return list or unconsolidated bins since
wdenk8bde7f72003-06-27 21:31:46 +00003295 no scheme using them consistently outperforms those that don't
3296 given above changes.
wdenk217c9da2002-10-25 20:35:49 +00003297 * Use best fit for very large chunks to prevent some worst-cases.
3298 * Added some support for debugging
3299
3300 V2.6.0 Sat Nov 4 07:05:23 1995 Doug Lea (dl at gee)
3301 * Removed footers when chunks are in use. Thanks to
wdenk8bde7f72003-06-27 21:31:46 +00003302 Paul Wilson (wilson@cs.texas.edu) for the suggestion.
wdenk217c9da2002-10-25 20:35:49 +00003303
3304 V2.5.4 Wed Nov 1 07:54:51 1995 Doug Lea (dl at gee)
3305 * Added malloc_trim, with help from Wolfram Gloger
wdenk8bde7f72003-06-27 21:31:46 +00003306 (wmglo@Dent.MED.Uni-Muenchen.DE).
wdenk217c9da2002-10-25 20:35:49 +00003307
3308 V2.5.3 Tue Apr 26 10:16:01 1994 Doug Lea (dl at g)
3309
3310 V2.5.2 Tue Apr 5 16:20:40 1994 Doug Lea (dl at g)
3311 * realloc: try to expand in both directions
3312 * malloc: swap order of clean-bin strategy;
3313 * realloc: only conditionally expand backwards
3314 * Try not to scavenge used bins
3315 * Use bin counts as a guide to preallocation
3316 * Occasionally bin return list chunks in first scan
3317 * Add a few optimizations from colin@nyx10.cs.du.edu
3318
3319 V2.5.1 Sat Aug 14 15:40:43 1993 Doug Lea (dl at g)
3320 * faster bin computation & slightly different binning
3321 * merged all consolidations to one part of malloc proper
wdenk8bde7f72003-06-27 21:31:46 +00003322 (eliminating old malloc_find_space & malloc_clean_bin)
wdenk217c9da2002-10-25 20:35:49 +00003323 * Scan 2 returns chunks (not just 1)
3324 * Propagate failure in realloc if malloc returns 0
3325 * Add stuff to allow compilation on non-ANSI compilers
wdenk8bde7f72003-06-27 21:31:46 +00003326 from kpv@research.att.com
wdenk217c9da2002-10-25 20:35:49 +00003327
3328 V2.5 Sat Aug 7 07:41:59 1993 Doug Lea (dl at g.oswego.edu)
3329 * removed potential for odd address access in prev_chunk
3330 * removed dependency on getpagesize.h
3331 * misc cosmetics and a bit more internal documentation
3332 * anticosmetics: mangled names in macros to evade debugger strangeness
3333 * tested on sparc, hp-700, dec-mips, rs6000
wdenk8bde7f72003-06-27 21:31:46 +00003334 with gcc & native cc (hp, dec only) allowing
3335 Detlefs & Zorn comparison study (in SIGPLAN Notices.)
wdenk217c9da2002-10-25 20:35:49 +00003336
3337 Trial version Fri Aug 28 13:14:29 1992 Doug Lea (dl at g.oswego.edu)
3338 * Based loosely on libg++-1.2X malloc. (It retains some of the overall
wdenk8bde7f72003-06-27 21:31:46 +00003339 structure of old version, but most details differ.)
wdenk217c9da2002-10-25 20:35:49 +00003340
3341*/