1 @comment !!! describe mmap et al (here?)
4 @node Memory Allocation, Character Handling, Error Reporting, Top
5 @chapter Memory Allocation
6 @cindex memory allocation
7 @cindex storage allocation
9 The GNU system provides several methods for allocating memory space
10 under explicit program control. They vary in generality and in
16 The @code{malloc} facility allows fully general dynamic allocation.
17 @xref{Unconstrained Allocation}.
20 Obstacks are another facility, less general than @code{malloc} but more
21 efficient and convenient for stacklike allocation. @xref{Obstacks}.
24 The function @code{alloca} lets you allocate storage dynamically that
25 will be freed automatically. @xref{Variable Size Automatic}.
30 * Memory Concepts:: An introduction to concepts and terminology.
31 * Dynamic Allocation and C:: How to get different kinds of allocation in C.
32 * Unconstrained Allocation:: The @code{malloc} facility allows fully general
34 * Allocation Debugging:: Finding memory leaks and not freed memory.
35 * Obstacks:: Obstacks are less general than malloc
36 but more efficient and convenient.
37 * Variable Size Automatic:: Allocation of variable-sized blocks
38 of automatic storage that are freed when the
39 calling function returns.
40 * Relocating Allocator:: Waste less memory, if you can tolerate
41 automatic relocation of the blocks you get.
45 @section Dynamic Memory Allocation Concepts
46 @cindex dynamic allocation
47 @cindex static allocation
48 @cindex automatic allocation
50 @dfn{Dynamic memory allocation} is a technique in which programs
51 determine as they are running where to store some information. You need
52 dynamic allocation when the number of memory blocks you need, or how
53 long you continue to need them, depends on the data you are working on.
55 For example, you may need a block to store a line read from an input file;
56 since there is no limit to how long a line can be, you must allocate the
57 storage dynamically and make it dynamically larger as you read more of the
60 Or, you may need a block for each record or each definition in the input
61 data; since you can't know in advance how many there will be, you must
62 allocate a new block for each record or definition as you read it.
64 When you use dynamic allocation, the allocation of a block of memory is an
65 action that the program requests explicitly. You call a function or macro
66 when you want to allocate space, and specify the size with an argument. If
67 you want to free the space, you do so by calling another function or macro.
68 You can do these things whenever you want, as often as you want.
70 @node Dynamic Allocation and C
71 @section Dynamic Allocation and C
73 The C language supports two kinds of memory allocation through the variables
78 @dfn{Static allocation} is what happens when you declare a static or
79 global variable. Each static or global variable defines one block of
80 space, of a fixed size. The space is allocated once, when your program
81 is started, and is never freed.
84 @dfn{Automatic allocation} happens when you declare an automatic
85 variable, such as a function argument or a local variable. The space
86 for an automatic variable is allocated when the compound statement
87 containing the declaration is entered, and is freed when that
88 compound statement is exited.
90 In GNU C, the length of the automatic storage can be an expression
91 that varies. In other C implementations, it must be a constant.
94 Dynamic allocation is not supported by C variables; there is no storage
95 class ``dynamic'', and there can never be a C variable whose value is
96 stored in dynamically allocated space. The only way to refer to
97 dynamically allocated space is through a pointer. Because it is less
98 convenient, and because the actual process of dynamic allocation
99 requires more computation time, programmers generally use dynamic
100 allocation only when neither static nor automatic allocation will serve.
102 For example, if you want to allocate dynamically some space to hold a
103 @code{struct foobar}, you cannot declare a variable of type @code{struct
104 foobar} whose contents are the dynamically allocated space. But you can
105 declare a variable of pointer type @code{struct foobar *} and assign it the
106 address of the space. Then you can use the operators @samp{*} and
107 @samp{->} on this pointer variable to refer to the contents of the space:
112 = (struct foobar *) malloc (sizeof (struct foobar));
114 ptr->next = current_foobar;
115 current_foobar = ptr;
119 @node Unconstrained Allocation
120 @section Unconstrained Allocation
121 @cindex unconstrained storage allocation
122 @cindex @code{malloc} function
123 @cindex heap, dynamic allocation from
125 The most general dynamic allocation facility is @code{malloc}. It
126 allows you to allocate blocks of memory of any size at any time, make
127 them bigger or smaller at any time, and free the blocks individually at
131 * Basic Allocation:: Simple use of @code{malloc}.
132 * Malloc Examples:: Examples of @code{malloc}. @code{xmalloc}.
133 * Freeing after Malloc:: Use @code{free} to free a block you
134 got with @code{malloc}.
135 * Changing Block Size:: Use @code{realloc} to make a block
137 * Allocating Cleared Space:: Use @code{calloc} to allocate a
139 * Efficiency and Malloc:: Efficiency considerations in use of
141 * Aligned Memory Blocks:: Allocating specially aligned memory:
142 @code{memalign} and @code{valloc}.
143 * Malloc Tunable Parameters:: Use @code{mallopt} to adjust allocation
145 * Heap Consistency Checking:: Automatic checking for errors.
146 * Hooks for Malloc:: You can use these hooks for debugging
147 programs that use @code{malloc}.
148 * Statistics of Malloc:: Getting information about how much
149 memory your program is using.
150 * Summary of Malloc:: Summary of @code{malloc} and related functions.
153 @node Basic Allocation
154 @subsection Basic Storage Allocation
155 @cindex allocation of memory with @code{malloc}
157 To allocate a block of memory, call @code{malloc}. The prototype for
158 this function is in @file{stdlib.h}.
161 @comment malloc.h stdlib.h
163 @deftypefun {void *} malloc (size_t @var{size})
164 This function returns a pointer to a newly allocated block @var{size}
165 bytes long, or a null pointer if the block could not be allocated.
168 The contents of the block are undefined; you must initialize it yourself
169 (or use @code{calloc} instead; @pxref{Allocating Cleared Space}).
170 Normally you would cast the value as a pointer to the kind of object
171 that you want to store in the block. Here we show an example of doing
172 so, and of initializing the space with zeros using the library function
173 @code{memset} (@pxref{Copying and Concatenation}):
178 ptr = (struct foo *) malloc (sizeof (struct foo));
179 if (ptr == 0) abort ();
180 memset (ptr, 0, sizeof (struct foo));
183 You can store the result of @code{malloc} into any pointer variable
184 without a cast, because @w{ISO C} automatically converts the type
185 @code{void *} to another type of pointer when necessary. But the cast
186 is necessary in contexts other than assignment operators or if you might
187 want your code to run in traditional C.
189 Remember that when allocating space for a string, the argument to
190 @code{malloc} must be one plus the length of the string. This is
191 because a string is terminated with a null character that doesn't count
192 in the ``length'' of the string but does need space. For example:
197 ptr = (char *) malloc (length + 1);
201 @xref{Representation of Strings}, for more information about this.
203 @node Malloc Examples
204 @subsection Examples of @code{malloc}
206 If no more space is available, @code{malloc} returns a null pointer.
207 You should check the value of @emph{every} call to @code{malloc}. It is
208 useful to write a subroutine that calls @code{malloc} and reports an
209 error if the value is a null pointer, returning only if the value is
210 nonzero. This function is conventionally called @code{xmalloc}. Here
215 xmalloc (size_t size)
217 register void *value = malloc (size);
219 fatal ("virtual memory exhausted");
224 Here is a real example of using @code{malloc} (by way of @code{xmalloc}).
225 The function @code{savestring} will copy a sequence of characters into
226 a newly allocated null-terminated string:
231 savestring (const char *ptr, size_t len)
233 register char *value = (char *) xmalloc (len + 1);
234 memcpy (value, ptr, len);
241 The block that @code{malloc} gives you is guaranteed to be aligned so
242 that it can hold any type of data. In the GNU system, the address is
243 always a multiple of eight on most systems, and a multiple of 16 on
244 64-bit systems. Only rarely is any higher boundary (such as a page
245 boundary) necessary; for those cases, use @code{memalign} or
246 @code{valloc} (@pxref{Aligned Memory Blocks}).
248 Note that the memory located after the end of the block is likely to be
249 in use for something else; perhaps a block already allocated by another
250 call to @code{malloc}. If you attempt to treat the block as longer than
251 you asked for it to be, you are liable to destroy the data that
252 @code{malloc} uses to keep track of its blocks, or you may destroy the
253 contents of another block. If you have already allocated a block and
254 discover you want it to be bigger, use @code{realloc} (@pxref{Changing
257 @node Freeing after Malloc
258 @subsection Freeing Memory Allocated with @code{malloc}
259 @cindex freeing memory allocated with @code{malloc}
260 @cindex heap, freeing memory from
262 When you no longer need a block that you got with @code{malloc}, use the
263 function @code{free} to make the block available to be allocated again.
264 The prototype for this function is in @file{stdlib.h}.
267 @comment malloc.h stdlib.h
269 @deftypefun void free (void *@var{ptr})
270 The @code{free} function deallocates the block of storage pointed at
276 @deftypefun void cfree (void *@var{ptr})
277 This function does the same thing as @code{free}. It's provided for
278 backward compatibility with SunOS; you should use @code{free} instead.
281 Freeing a block alters the contents of the block. @strong{Do not expect to
282 find any data (such as a pointer to the next block in a chain of blocks) in
283 the block after freeing it.} Copy whatever you need out of the block before
284 freeing it! Here is an example of the proper way to free all the blocks in
285 a chain, and the strings that they point to:
295 free_chain (struct chain *chain)
299 struct chain *next = chain->next;
307 Occasionally, @code{free} can actually return memory to the operating
308 system and make the process smaller. Usually, all it can do is allow a
309 later call to @code{malloc} to reuse the space. In the meantime, the
310 space remains in your program as part of a free-list used internally by
313 There is no point in freeing blocks at the end of a program, because all
314 of the program's space is given back to the system when the process
317 @node Changing Block Size
318 @subsection Changing the Size of a Block
319 @cindex changing the size of a block (@code{malloc})
321 Often you do not know for certain how big a block you will ultimately need
322 at the time you must begin to use the block. For example, the block might
323 be a buffer that you use to hold a line being read from a file; no matter
324 how long you make the buffer initially, you may encounter a line that is
327 You can make the block longer by calling @code{realloc}. This function
328 is declared in @file{stdlib.h}.
331 @comment malloc.h stdlib.h
333 @deftypefun {void *} realloc (void *@var{ptr}, size_t @var{newsize})
334 The @code{realloc} function changes the size of the block whose address is
335 @var{ptr} to be @var{newsize}.
337 Since the space after the end of the block may be in use, @code{realloc}
338 may find it necessary to copy the block to a new address where more free
339 space is available. The value of @code{realloc} is the new address of the
340 block. If the block needs to be moved, @code{realloc} copies the old
343 If you pass a null pointer for @var{ptr}, @code{realloc} behaves just
344 like @samp{malloc (@var{newsize})}. This can be convenient, but beware
345 that older implementations (before @w{ISO C}) may not support this
346 behavior, and will probably crash when @code{realloc} is passed a null
350 Like @code{malloc}, @code{realloc} may return a null pointer if no
351 memory space is available to make the block bigger. When this happens,
352 the original block is untouched; it has not been modified or relocated.
354 In most cases it makes no difference what happens to the original block
355 when @code{realloc} fails, because the application program cannot continue
356 when it is out of memory, and the only thing to do is to give a fatal error
357 message. Often it is convenient to write and use a subroutine,
358 conventionally called @code{xrealloc}, that takes care of the error message
359 as @code{xmalloc} does for @code{malloc}:
363 xrealloc (void *ptr, size_t size)
365 register void *value = realloc (ptr, size);
367 fatal ("Virtual memory exhausted");
372 You can also use @code{realloc} to make a block smaller. The reason you
374 @comment The following is no longer true with the new malloc.
375 @comment But it seems wise to keep the warning for other implementations.
376 In several allocation implementations, making a block smaller sometimes
377 necessitates copying it, so it can fail if no other space is available.
379 If the new size you specify is the same as the old size, @code{realloc}
380 is guaranteed to change nothing and return the same address that you gave.
382 @node Allocating Cleared Space
383 @subsection Allocating Cleared Space
385 The function @code{calloc} allocates memory and clears it to zero. It
386 is declared in @file{stdlib.h}.
389 @comment malloc.h stdlib.h
391 @deftypefun {void *} calloc (size_t @var{count}, size_t @var{eltsize})
392 This function allocates a block long enough to contain a vector of
393 @var{count} elements, each of size @var{eltsize}. Its contents are
394 cleared to zero before @code{calloc} returns.
397 You could define @code{calloc} as follows:
401 calloc (size_t count, size_t eltsize)
403 size_t size = count * eltsize;
404 void *value = malloc (size);
406 memset (value, 0, size);
411 But in general, it is not guaranteed that @code{calloc} calls
412 @code{malloc} internally. Therefore, if an application provides its own
413 @code{malloc}/@code{realloc}/@code{free} outside the C library, it
414 should always define @code{calloc}, too.
416 @node Efficiency and Malloc
417 @subsection Efficiency Considerations for @code{malloc}
418 @cindex efficiency and @code{malloc}
422 @c No longer true, see below instead.
423 To make the best use of @code{malloc}, it helps to know that the GNU
424 version of @code{malloc} always dispenses small amounts of memory in
425 blocks whose sizes are powers of two. It keeps separate pools for each
426 power of two. This holds for sizes up to a page size. Therefore, if
427 you are free to choose the size of a small block in order to make
428 @code{malloc} more efficient, make it a power of two.
429 @c !!! xref getpagesize
431 Once a page is split up for a particular block size, it can't be reused
432 for another size unless all the blocks in it are freed. In many
433 programs, this is unlikely to happen. Thus, you can sometimes make a
434 program use memory more efficiently by using blocks of the same size for
435 many different purposes.
437 When you ask for memory blocks of a page or larger, @code{malloc} uses a
438 different strategy; it rounds the size up to a multiple of a page, and
439 it can coalesce and split blocks as needed.
441 The reason for the two strategies is that it is important to allocate
442 and free small blocks as fast as possible, but speed is less important
443 for a large block since the program normally spends a fair amount of
444 time using it. Also, large blocks are normally fewer in number.
445 Therefore, for large blocks, it makes sense to use a method which takes
446 more time to minimize the wasted space.
450 As apposed to other versions, the @code{malloc} in GNU libc does not
451 round up block sizes to powers of two, neither for large nor for small
452 sizes. Neighboring chunks can be coalesced on a @code{free} no matter
453 what their size is. This makes the implementation suitable for all
454 kinds of allocation patterns without generally incurring high memory
455 waste through fragmentation.
457 Very large blocks (much larger than a page) are allocated with
458 @code{mmap} (anonymous or via @code{/dev/zero}) by this implementation.
459 This has the great advantage that these chunks are returned to the
460 system immediately when they are freed. Therefore, it cannot happen
461 that a large chunk becomes ``locked'' in between smaller ones and even
462 after calling @code{free} wastes memory. The size threshold for
463 @code{mmap} to be used can be adjusted with @code{mallopt}. The use of
464 @code{mmap} can also be disabled completely.
466 @node Aligned Memory Blocks
467 @subsection Allocating Aligned Memory Blocks
469 @cindex page boundary
470 @cindex alignment (with @code{malloc})
472 The address of a block returned by @code{malloc} or @code{realloc} in
473 the GNU system is always a multiple of eight (or sixteen on 64-bit
474 systems). If you need a block whose address is a multiple of a higher
475 power of two than that, use @code{memalign} or @code{valloc}. These
476 functions are declared in @file{stdlib.h}.
478 With the GNU library, you can use @code{free} to free the blocks that
479 @code{memalign} and @code{valloc} return. That does not work in BSD,
480 however---BSD does not provide any way to free such blocks.
482 @comment malloc.h stdlib.h
484 @deftypefun {void *} memalign (size_t @var{boundary}, size_t @var{size})
485 The @code{memalign} function allocates a block of @var{size} bytes whose
486 address is a multiple of @var{boundary}. The @var{boundary} must be a
487 power of two! The function @code{memalign} works by allocating a
488 somewhat larger block, and then returning an address within the block
489 that is on the specified boundary.
492 @comment malloc.h stdlib.h
494 @deftypefun {void *} valloc (size_t @var{size})
495 Using @code{valloc} is like using @code{memalign} and passing the page size
496 as the value of the second argument. It is implemented like this:
502 return memalign (getpagesize (), size);
505 @c !!! xref getpagesize
508 @node Malloc Tunable Parameters
509 @subsection Malloc Tunable Parameters
511 You can adjust some parameters for dynamic memory allocation with the
512 @code{mallopt} function. This function is the general SVID/XPG
513 interface, defined in @file{malloc.h}.
516 @deftypefun int mallopt (int @var{param}, int @var{value})
517 When calling @code{mallopt}, the @var{param} argument specifies the
518 parameter to be set, and @var{value} the new value to be set. Possible
519 choices for @var{param}, as defined in @file{malloc.h}, are:
522 @item M_TRIM_THRESHOLD
523 This is the minimum size (in bytes) of the top-most, releaseable chunk
524 that will cause @code{sbrk} to be called with a negative argument in
525 order to return memory to the system.
527 This parameter determines the amount of extra memory to obtain from the
528 system when a call to @code{sbrk} is required. It also specifies the
529 number of bytes to retain when shrinking the heap by calling @code{sbrk}
530 with a negative argument. This provides the necessary hysteresis in
531 heap size such that excessive amounts of system calls can be avoided.
532 @item M_MMAP_THRESHOLD
533 All chunks larger than this value are allocated outside the normal
534 heap, using the @code{mmap} system call. This way it is guaranteed
535 that the memory for these chunks can be returned to the system on
538 The maximum number of chunks to allocate with @code{mmap}. Setting this
539 to zero disables all use of @code{mmap}.
544 @node Heap Consistency Checking
545 @subsection Heap Consistency Checking
547 @cindex heap consistency checking
548 @cindex consistency checking, of heap
550 You can ask @code{malloc} to check the consistency of dynamic storage by
551 using the @code{mcheck} function. This function is a GNU extension,
552 declared in @file{malloc.h}.
557 @deftypefun int mcheck (void (*@var{abortfn}) (enum mcheck_status @var{status}))
558 Calling @code{mcheck} tells @code{malloc} to perform occasional
559 consistency checks. These will catch things such as writing
560 past the end of a block that was allocated with @code{malloc}.
562 The @var{abortfn} argument is the function to call when an inconsistency
563 is found. If you supply a null pointer, then @code{mcheck} uses a
564 default function which prints a message and calls @code{abort}
565 (@pxref{Aborting a Program}). The function you supply is called with
566 one argument, which says what sort of inconsistency was detected; its
567 type is described below.
569 It is too late to begin allocation checking once you have allocated
570 anything with @code{malloc}. So @code{mcheck} does nothing in that
571 case. The function returns @code{-1} if you call it too late, and
572 @code{0} otherwise (when it is successful).
574 The easiest way to arrange to call @code{mcheck} early enough is to use
575 the option @samp{-lmcheck} when you link your program; then you don't
576 need to modify your program source at all.
579 @deftypefun {enum mcheck_status} mprobe (void *@var{pointer})
580 The @code{mprobe} function lets you explicitly check for inconsistencies
581 in a particular allocated block. You must have already called
582 @code{mcheck} at the beginning of the program, to do its occasional
583 checks; calling @code{mprobe} requests an additional consistency check
584 to be done at the time of the call.
586 The argument @var{pointer} must be a pointer returned by @code{malloc}
587 or @code{realloc}. @code{mprobe} returns a value that says what
588 inconsistency, if any, was found. The values are described below.
591 @deftp {Data Type} {enum mcheck_status}
592 This enumerated type describes what kind of inconsistency was detected
593 in an allocated block, if any. Here are the possible values:
596 @item MCHECK_DISABLED
597 @code{mcheck} was not called before the first allocation.
598 No consistency checking can be done.
600 No inconsistency detected.
602 The data immediately before the block was modified.
603 This commonly happens when an array index or pointer
604 is decremented too far.
606 The data immediately after the block was modified.
607 This commonly happens when an array index or pointer
608 is incremented too far.
610 The block was already freed.
614 @node Hooks for Malloc
615 @subsection Storage Allocation Hooks
616 @cindex allocation hooks, for @code{malloc}
618 The GNU C library lets you modify the behavior of @code{malloc},
619 @code{realloc}, and @code{free} by specifying appropriate hook
620 functions. You can use these hooks to help you debug programs that use
621 dynamic storage allocation, for example.
623 The hook variables are declared in @file{malloc.h}.
628 @defvar __malloc_hook
629 The value of this variable is a pointer to function that @code{malloc}
630 uses whenever it is called. You should define this function to look
631 like @code{malloc}; that is, like:
634 void *@var{function} (size_t @var{size}, void *@var{caller})
637 The value of @var{caller} is the return address found on the stack when
638 the @code{malloc} function was called. This value allows to trace the
639 memory consumption of the program.
644 @defvar __realloc_hook
645 The value of this variable is a pointer to function that @code{realloc}
646 uses whenever it is called. You should define this function to look
647 like @code{realloc}; that is, like:
650 void *@var{function} (void *@var{ptr}, size_t @var{size}, void *@var{caller})
653 The value of @var{caller} is the return address found on the stack when
654 the @code{realloc} function was called. This value allows to trace the
655 memory consumption of the program.
661 The value of this variable is a pointer to function that @code{free}
662 uses whenever it is called. You should define this function to look
663 like @code{free}; that is, like:
666 void @var{function} (void *@var{ptr}, void *@var{caller})
669 The value of @var{caller} is the return address found on the stack when
670 the @code{free} function was called. This value allows to trace the
671 memory consumption of the program.
674 You must make sure that the function you install as a hook for one of
675 these functions does not call that function recursively without restoring
676 the old value of the hook first! Otherwise, your program will get stuck
677 in an infinite recursion.
679 Here is an example showing how to use @code{__malloc_hook} properly. It
680 installs a function that prints out information every time @code{malloc}
684 static void *(*old_malloc_hook) (size_t);
686 my_malloc_hook (size_t size)
689 __malloc_hook = old_malloc_hook;
690 result = malloc (size);
691 /* @r{@code{printf} might call @code{malloc}, so protect it too.} */
692 printf ("malloc (%u) returns %p\n", (unsigned int) size, result);
693 __malloc_hook = my_malloc_hook;
700 old_malloc_hook = __malloc_hook;
701 __malloc_hook = my_malloc_hook;
706 The @code{mcheck} function (@pxref{Heap Consistency Checking}) works by
707 installing such hooks.
709 @c __morecore, __after_morecore_hook are undocumented
710 @c It's not clear whether to document them.
712 @node Statistics of Malloc
713 @subsection Statistics for Storage Allocation with @code{malloc}
715 @cindex allocation statistics
716 You can get information about dynamic storage allocation by calling the
717 @code{mallinfo} function. This function and its associated data type
718 are declared in @file{malloc.h}; they are an extension of the standard
724 @deftp {Data Type} {struct mallinfo}
725 This structure type is used to return information about the dynamic
726 storage allocator. It contains the following members:
730 This is the total size of memory allocated with @code{sbrk} by
731 @code{malloc}, in bytes.
734 This is the number of chunks not in use. (The storage allocator
735 internally gets chunks of memory from the operating system, and then
736 carves them up to satisfy individual @code{malloc} requests; see
737 @ref{Efficiency and Malloc}.)
740 This field is unused.
743 This is the total number of chunks allocated with @code{mmap}.
746 This is the total size of memory allocated with @code{mmap}, in bytes.
749 This field is unused.
752 This field is unused.
755 This is the total size of memory occupied by chunks handed out by
759 This is the total size of memory occupied by free (not in use) chunks.
762 This is the size of the top-most, releaseable chunk that normally
763 borders the end of the heap (i.e. the ``brk'' of the process).
770 @deftypefun {struct mallinfo} mallinfo (void)
771 This function returns information about the current dynamic memory usage
772 in a structure of type @code{struct mallinfo}.
775 @node Summary of Malloc
776 @subsection Summary of @code{malloc}-Related Functions
778 Here is a summary of the functions that work with @code{malloc}:
781 @item void *malloc (size_t @var{size})
782 Allocate a block of @var{size} bytes. @xref{Basic Allocation}.
784 @item void free (void *@var{addr})
785 Free a block previously allocated by @code{malloc}. @xref{Freeing after
788 @item void *realloc (void *@var{addr}, size_t @var{size})
789 Make a block previously allocated by @code{malloc} larger or smaller,
790 possibly by copying it to a new location. @xref{Changing Block Size}.
792 @item void *calloc (size_t @var{count}, size_t @var{eltsize})
793 Allocate a block of @var{count} * @var{eltsize} bytes using
794 @code{malloc}, and set its contents to zero. @xref{Allocating Cleared
797 @item void *valloc (size_t @var{size})
798 Allocate a block of @var{size} bytes, starting on a page boundary.
799 @xref{Aligned Memory Blocks}.
801 @item void *memalign (size_t @var{size}, size_t @var{boundary})
802 Allocate a block of @var{size} bytes, starting on an address that is a
803 multiple of @var{boundary}. @xref{Aligned Memory Blocks}.
805 @item int mallopt (int @var{param}, int @var{value})
806 Adjust a tunable parameter. @xref{Malloc Tunable Parameters}
808 @item int mcheck (void (*@var{abortfn}) (void))
809 Tell @code{malloc} to perform occasional consistency checks on
810 dynamically allocated memory, and to call @var{abortfn} when an
811 inconsistency is found. @xref{Heap Consistency Checking}.
813 @item void *(*__malloc_hook) (size_t @var{size}, void *@var{caller})
814 A pointer to a function that @code{malloc} uses whenever it is called.
816 @item void *(*__realloc_hook) (void *@var{ptr}, size_t @var{size}, void *@var{caller})
817 A pointer to a function that @code{realloc} uses whenever it is called.
819 @item void (*__free_hook) (void *@var{ptr}, void *@var{caller})
820 A pointer to a function that @code{free} uses whenever it is called.
822 @item struct mallinfo mallinfo (void)
823 Return information about the current dynamic memory usage.
824 @xref{Statistics of Malloc}.
827 @node Allocation Debugging
828 @section Allocation Debugging
829 @cindex allocation debugging
830 @cindex malloc debugger
832 An complicated task when programming with languages which do not use
833 garbage collected dynamic memory allocation is to find memory leaks.
834 Long running programs must assure that dynamically allocated objects are
835 freed at the end of their lifetime. If this does not happen the system
836 runs out of memory, sooner or later.
838 The @code{malloc} implementation in the GNU C library provides some
839 simple means to detect sich leaks and provide some information to find
840 the location. To do this the application must be started in a special
841 mode which is enabled by an environment variable. There are no speed
842 penalties if the program is compiled in preparation of the debugging if
843 the debug mode is not enabled.
846 * Tracing malloc:: How to install the tracing functionality.
847 * Using the Memory Debugger:: Example programs excerpts.
848 * Tips for the Memory Debugger:: Some more or less clever ideas.
849 * Interpreting the traces:: What do all these lines mean?
853 @subsection How to install the tracing functionality
857 @deftypefun void mtrace (void)
858 When the @code{mtrace} function is called it looks for an environment
859 variable named @code{MALLOC_TRACE}. This variable is supposed to
860 contain a valid file name. The user must have write access. If the
861 file already exists it is truncated. If the environment variable is not
862 set or it does not name a valid file which can be opened for writing
863 nothing is done. The behaviour of @code{malloc} etc. is not changed.
864 For obvious reasons this also happens if the application is install SUID
867 If the named file is successfully opened @code{mtrace} installs special
868 handlers for the functions @code{malloc}, @code{realloc}, and
869 @code{free} (@pxref{Hooks for Malloc}). From now on all uses of these
870 functions are traced and protocolled into the file. There is now of
871 course a speed penalty for all calls to the traced functions so that the
872 tracing should not be enabled during their normal use.
874 This function is a GNU extension and generally not available on other
875 systems. The prototype can be found in @file{mcheck.h}.
880 @deftypefun void muntrace (void)
881 The @code{muntrace} function can be called after @code{mtrace} was used
882 to enable tracing the @code{malloc} calls. If no (succesful) call of
883 @code{mtrace} was made @code{muntrace} does nothing.
885 Otherwise it deinstalls the handlers for @code{malloc}, @code{realloc},
886 and @code{free} and then closes the protocol file. No calls are
887 protocolled anymore and the programs runs again with the full speed.
889 This function is a GNU extension and generally not available on other
890 systems. The prototype can be found in @file{mcheck.h}.
893 @node Using the Memory Debugger
894 @subsection Example programs excerpts
896 Even though the tracing functionality does not influence the runtime
897 behaviour of the program it is no wise idea to call @code{mtrace} in all
898 programs. Just imagine you debug a program using @code{mtrace} and all
899 other programs used in the debug sessions also trace their @code{malloc}
900 calls. The output file would be the same for all programs and so is
901 unusable. Therefore on should call @code{mtrace} only if compiled for
902 debugging. A program could therefore start like this:
908 main (int argc, char *argv[])
917 This is all what is needed if you want to trace the calls during the
918 whole runtime of the program. Alternatively you can stop the tracing at
919 any time with a call to @code{muntrace}. It is even possible to restart
920 the tracing again with a new call to @code{mtrace}. But this can course
921 unreliable results since there are possibly calls of the functions which
922 are not called. Please note that not only the application uses the
923 traced functions, also libraries (including the C library itself) use
926 This last point is also why it is no good idea to call @code{muntrace}
927 before the program terminated. The libraries are informed about the
928 termination of the program only after the program returns from
929 @code{main} or calls @code{exit} and so cannot free the memory they use
932 So the best thing one can do is to call @code{mtrace} as the very first
933 function in the program and never call @code{muntrace}. So the program
934 traces almost all uses of the @code{malloc} functions (except those
935 calls which are executed by constructors of the program or used
938 @node Tips for the Memory Debugger
939 @subsection Some more or less clever ideas
941 You know the situation. The program is prepared for debugging and in
942 all debugging sessions it runs well. But once it is started without
943 debugging the error shows up. In our situation here: the memory leaks
944 becomes visible only when we just turned off the debugging. If you
945 foresee such situations you can still win. Simply use something
946 equivalent to the following little program:
956 signal (SIGUSR1, enable);
963 signal (SIGUSR2, disable);
967 main (int argc, char *argv[])
971 signal (SIGUSR1, enable);
972 signal (SIGUSR2, disable);
978 I.e., the user can start the memory debugger any time he wants if the
979 program was started with @code{MALLOC_TRACE} set in the environment.
980 The output will of course not show the allocations which happened before
981 the first signal but if there is a memory leak this will show up
984 @node Interpreting the traces
985 @subsection Interpreting the traces
987 If you take a look at the output it will look similar to this:
991 @ [0x8048209] - 0x8064cc8
992 @ [0x8048209] - 0x8064ce0
993 @ [0x8048209] - 0x8064cf8
994 @ [0x80481eb] + 0x8064c48 0x14
995 @ [0x80481eb] + 0x8064c60 0x14
996 @ [0x80481eb] + 0x8064c78 0x14
997 @ [0x80481eb] + 0x8064c90 0x14
1001 What this all means is not really important since the trace file is not
1002 meant to be read by a human. Therefore no attention is payed to good
1003 readability. Instead there is a program which comes with the GNU C
1004 library which interprets the traces and outputs a summary in on
1005 user-friendly way. The program is called @code{mtrace} (it is in fact a
1006 Perl script) and it takes one or two arguments. In any case the name of
1007 the file with the trace must be specified. If an optional precedes the
1008 name of the trace file this must be the name of the program which
1009 generated the trace.
1012 drepper$ mtrace tst-mtrace log
1016 In this case the program @code{tst-mtrace} was run and it produced a
1017 trace file @file{log}. The message printed by @code{mtrace} shows there
1018 are no problems with the code, all allocated memory was freed
1021 If we call @code{mtrace} on the example trace given above we would get a
1025 drepper$ mtrace errlog
1026 - 0x08064cc8 Free 2 was never alloc'd 0x8048209
1027 - 0x08064ce0 Free 3 was never alloc'd 0x8048209
1028 - 0x08064cf8 Free 4 was never alloc'd 0x8048209
1033 0x08064c48 0x14 at 0x80481eb
1034 0x08064c60 0x14 at 0x80481eb
1035 0x08064c78 0x14 at 0x80481eb
1036 0x08064c90 0x14 at 0x80481eb
1039 We have called @code{mtrace} with only one argument and so the script
1040 has no chance to find out what is meant with the addresses given in the
1041 trace. We can do better:
1044 drepper$ mtrace tst-mtrace errlog
1045 - 0x08064cc8 Free 2 was never alloc'd /home/drepper/tst-mtrace.c:39
1046 - 0x08064ce0 Free 3 was never alloc'd /home/drepper/tst-mtrace.c:39
1047 - 0x08064cf8 Free 4 was never alloc'd /home/drepper/tst-mtrace.c:39
1052 0x08064c48 0x14 at /home/drepper/tst-mtrace.c:33
1053 0x08064c60 0x14 at /home/drepper/tst-mtrace.c:33
1054 0x08064c78 0x14 at /home/drepper/tst-mtrace.c:33
1055 0x08064c90 0x14 at /home/drepper/tst-mtrace.c:33
1058 Suddenly the output makes much more sense and the user can see
1059 immediately where the function calls causing the trouble can be found.
1065 An @dfn{obstack} is a pool of memory containing a stack of objects. You
1066 can create any number of separate obstacks, and then allocate objects in
1067 specified obstacks. Within each obstack, the last object allocated must
1068 always be the first one freed, but distinct obstacks are independent of
1071 Aside from this one constraint of order of freeing, obstacks are totally
1072 general: an obstack can contain any number of objects of any size. They
1073 are implemented with macros, so allocation is usually very fast as long as
1074 the objects are usually small. And the only space overhead per object is
1075 the padding needed to start each object on a suitable boundary.
1078 * Creating Obstacks:: How to declare an obstack in your program.
1079 * Preparing for Obstacks:: Preparations needed before you can
1081 * Allocation in an Obstack:: Allocating objects in an obstack.
1082 * Freeing Obstack Objects:: Freeing objects in an obstack.
1083 * Obstack Functions:: The obstack functions are both
1084 functions and macros.
1085 * Growing Objects:: Making an object bigger by stages.
1086 * Extra Fast Growing:: Extra-high-efficiency (though more
1087 complicated) growing objects.
1088 * Status of an Obstack:: Inquiries about the status of an obstack.
1089 * Obstacks Data Alignment:: Controlling alignment of objects in obstacks.
1090 * Obstack Chunks:: How obstacks obtain and release chunks;
1091 efficiency considerations.
1092 * Summary of Obstacks::
1095 @node Creating Obstacks
1096 @subsection Creating Obstacks
1098 The utilities for manipulating obstacks are declared in the header
1099 file @file{obstack.h}.
1104 @deftp {Data Type} {struct obstack}
1105 An obstack is represented by a data structure of type @code{struct
1106 obstack}. This structure has a small fixed size; it records the status
1107 of the obstack and how to find the space in which objects are allocated.
1108 It does not contain any of the objects themselves. You should not try
1109 to access the contents of the structure directly; use only the functions
1110 described in this chapter.
1113 You can declare variables of type @code{struct obstack} and use them as
1114 obstacks, or you can allocate obstacks dynamically like any other kind
1115 of object. Dynamic allocation of obstacks allows your program to have a
1116 variable number of different stacks. (You can even allocate an
1117 obstack structure in another obstack, but this is rarely useful.)
1119 All the functions that work with obstacks require you to specify which
1120 obstack to use. You do this with a pointer of type @code{struct obstack
1121 *}. In the following, we often say ``an obstack'' when strictly
1122 speaking the object at hand is such a pointer.
1124 The objects in the obstack are packed into large blocks called
1125 @dfn{chunks}. The @code{struct obstack} structure points to a chain of
1126 the chunks currently in use.
1128 The obstack library obtains a new chunk whenever you allocate an object
1129 that won't fit in the previous chunk. Since the obstack library manages
1130 chunks automatically, you don't need to pay much attention to them, but
1131 you do need to supply a function which the obstack library should use to
1132 get a chunk. Usually you supply a function which uses @code{malloc}
1133 directly or indirectly. You must also supply a function to free a chunk.
1134 These matters are described in the following section.
1136 @node Preparing for Obstacks
1137 @subsection Preparing for Using Obstacks
1139 Each source file in which you plan to use the obstack functions
1140 must include the header file @file{obstack.h}, like this:
1143 #include <obstack.h>
1146 @findex obstack_chunk_alloc
1147 @findex obstack_chunk_free
1148 Also, if the source file uses the macro @code{obstack_init}, it must
1149 declare or define two functions or macros that will be called by the
1150 obstack library. One, @code{obstack_chunk_alloc}, is used to allocate
1151 the chunks of memory into which objects are packed. The other,
1152 @code{obstack_chunk_free}, is used to return chunks when the objects in
1153 them are freed. These macros should appear before any use of obstacks
1156 Usually these are defined to use @code{malloc} via the intermediary
1157 @code{xmalloc} (@pxref{Unconstrained Allocation}). This is done with
1158 the following pair of macro definitions:
1161 #define obstack_chunk_alloc xmalloc
1162 #define obstack_chunk_free free
1166 Though the storage you get using obstacks really comes from @code{malloc},
1167 using obstacks is faster because @code{malloc} is called less often, for
1168 larger blocks of memory. @xref{Obstack Chunks}, for full details.
1170 At run time, before the program can use a @code{struct obstack} object
1171 as an obstack, it must initialize the obstack by calling
1172 @code{obstack_init}.
1176 @deftypefun int obstack_init (struct obstack *@var{obstack-ptr})
1177 Initialize obstack @var{obstack-ptr} for allocation of objects. This
1178 function calls the obstack's @code{obstack_chunk_alloc} function. It
1179 returns 0 if @code{obstack_chunk_alloc} returns a null pointer, meaning
1180 that it is out of memory. Otherwise, it returns 1. If you supply an
1181 @code{obstack_chunk_alloc} function that calls @code{exit}
1182 (@pxref{Program Termination}) or @code{longjmp} (@pxref{Non-Local
1183 Exits}) when out of memory, you can safely ignore the value that
1184 @code{obstack_init} returns.
1187 Here are two examples of how to allocate the space for an obstack and
1188 initialize it. First, an obstack that is a static variable:
1191 static struct obstack myobstack;
1193 obstack_init (&myobstack);
1197 Second, an obstack that is itself dynamically allocated:
1200 struct obstack *myobstack_ptr
1201 = (struct obstack *) xmalloc (sizeof (struct obstack));
1203 obstack_init (myobstack_ptr);
1206 @node Allocation in an Obstack
1207 @subsection Allocation in an Obstack
1208 @cindex allocation (obstacks)
1210 The most direct way to allocate an object in an obstack is with
1211 @code{obstack_alloc}, which is invoked almost like @code{malloc}.
1215 @deftypefun {void *} obstack_alloc (struct obstack *@var{obstack-ptr}, int @var{size})
1216 This allocates an uninitialized block of @var{size} bytes in an obstack
1217 and returns its address. Here @var{obstack-ptr} specifies which obstack
1218 to allocate the block in; it is the address of the @code{struct obstack}
1219 object which represents the obstack. Each obstack function or macro
1220 requires you to specify an @var{obstack-ptr} as the first argument.
1222 This function calls the obstack's @code{obstack_chunk_alloc} function if
1223 it needs to allocate a new chunk of memory; it returns a null pointer if
1224 @code{obstack_chunk_alloc} returns one. In that case, it has not
1225 changed the amount of memory allocated in the obstack. If you supply an
1226 @code{obstack_chunk_alloc} function that calls @code{exit}
1227 (@pxref{Program Termination}) or @code{longjmp} (@pxref{Non-Local
1228 Exits}) when out of memory, then @code{obstack_alloc} will never return
1232 For example, here is a function that allocates a copy of a string @var{str}
1233 in a specific obstack, which is in the variable @code{string_obstack}:
1236 struct obstack string_obstack;
1239 copystring (char *string)
1241 size_t len = strlen (string) + 1;
1242 char *s = (char *) obstack_alloc (&string_obstack, len);
1243 memcpy (s, string, len);
1248 To allocate a block with specified contents, use the function
1249 @code{obstack_copy}, declared like this:
1253 @deftypefun {void *} obstack_copy (struct obstack *@var{obstack-ptr}, void *@var{address}, int @var{size})
1254 This allocates a block and initializes it by copying @var{size}
1255 bytes of data starting at @var{address}. It can return a null pointer
1256 under the same conditions as @code{obstack_alloc}.
1261 @deftypefun {void *} obstack_copy0 (struct obstack *@var{obstack-ptr}, void *@var{address}, int @var{size})
1262 Like @code{obstack_copy}, but appends an extra byte containing a null
1263 character. This extra byte is not counted in the argument @var{size}.
1266 The @code{obstack_copy0} function is convenient for copying a sequence
1267 of characters into an obstack as a null-terminated string. Here is an
1272 obstack_savestring (char *addr, int size)
1274 return obstack_copy0 (&myobstack, addr, size);
1279 Contrast this with the previous example of @code{savestring} using
1280 @code{malloc} (@pxref{Basic Allocation}).
1282 @node Freeing Obstack Objects
1283 @subsection Freeing Objects in an Obstack
1284 @cindex freeing (obstacks)
1286 To free an object allocated in an obstack, use the function
1287 @code{obstack_free}. Since the obstack is a stack of objects, freeing
1288 one object automatically frees all other objects allocated more recently
1289 in the same obstack.
1293 @deftypefun void obstack_free (struct obstack *@var{obstack-ptr}, void *@var{object})
1294 If @var{object} is a null pointer, everything allocated in the obstack
1295 is freed. Otherwise, @var{object} must be the address of an object
1296 allocated in the obstack. Then @var{object} is freed, along with
1297 everything allocated in @var{obstack} since @var{object}.
1300 Note that if @var{object} is a null pointer, the result is an
1301 uninitialized obstack. To free all storage in an obstack but leave it
1302 valid for further allocation, call @code{obstack_free} with the address
1303 of the first object allocated on the obstack:
1306 obstack_free (obstack_ptr, first_object_allocated_ptr);
1309 Recall that the objects in an obstack are grouped into chunks. When all
1310 the objects in a chunk become free, the obstack library automatically
1311 frees the chunk (@pxref{Preparing for Obstacks}). Then other
1312 obstacks, or non-obstack allocation, can reuse the space of the chunk.
1314 @node Obstack Functions
1315 @subsection Obstack Functions and Macros
1318 The interfaces for using obstacks may be defined either as functions or
1319 as macros, depending on the compiler. The obstack facility works with
1320 all C compilers, including both @w{ISO C} and traditional C, but there are
1321 precautions you must take if you plan to use compilers other than GNU C.
1323 If you are using an old-fashioned @w{non-ISO C} compiler, all the obstack
1324 ``functions'' are actually defined only as macros. You can call these
1325 macros like functions, but you cannot use them in any other way (for
1326 example, you cannot take their address).
1328 Calling the macros requires a special precaution: namely, the first
1329 operand (the obstack pointer) may not contain any side effects, because
1330 it may be computed more than once. For example, if you write this:
1333 obstack_alloc (get_obstack (), 4);
1337 you will find that @code{get_obstack} may be called several times.
1338 If you use @code{*obstack_list_ptr++} as the obstack pointer argument,
1339 you will get very strange results since the incrementation may occur
1342 In @w{ISO C}, each function has both a macro definition and a function
1343 definition. The function definition is used if you take the address of the
1344 function without calling it. An ordinary call uses the macro definition by
1345 default, but you can request the function definition instead by writing the
1346 function name in parentheses, as shown here:
1351 /* @r{Use the macro}. */
1352 x = (char *) obstack_alloc (obptr, size);
1353 /* @r{Call the function}. */
1354 x = (char *) (obstack_alloc) (obptr, size);
1355 /* @r{Take the address of the function}. */
1356 funcp = obstack_alloc;
1360 This is the same situation that exists in @w{ISO C} for the standard library
1361 functions. @xref{Macro Definitions}.
1363 @strong{Warning:} When you do use the macros, you must observe the
1364 precaution of avoiding side effects in the first operand, even in @w{ISO C}.
1366 If you use the GNU C compiler, this precaution is not necessary, because
1367 various language extensions in GNU C permit defining the macros so as to
1368 compute each argument only once.
1370 @node Growing Objects
1371 @subsection Growing Objects
1372 @cindex growing objects (in obstacks)
1373 @cindex changing the size of a block (obstacks)
1375 Because storage in obstack chunks is used sequentially, it is possible to
1376 build up an object step by step, adding one or more bytes at a time to the
1377 end of the object. With this technique, you do not need to know how much
1378 data you will put in the object until you come to the end of it. We call
1379 this the technique of @dfn{growing objects}. The special functions
1380 for adding data to the growing object are described in this section.
1382 You don't need to do anything special when you start to grow an object.
1383 Using one of the functions to add data to the object automatically
1384 starts it. However, it is necessary to say explicitly when the object is
1385 finished. This is done with the function @code{obstack_finish}.
1387 The actual address of the object thus built up is not known until the
1388 object is finished. Until then, it always remains possible that you will
1389 add so much data that the object must be copied into a new chunk.
1391 While the obstack is in use for a growing object, you cannot use it for
1392 ordinary allocation of another object. If you try to do so, the space
1393 already added to the growing object will become part of the other object.
1397 @deftypefun void obstack_blank (struct obstack *@var{obstack-ptr}, int @var{size})
1398 The most basic function for adding to a growing object is
1399 @code{obstack_blank}, which adds space without initializing it.
1404 @deftypefun void obstack_grow (struct obstack *@var{obstack-ptr}, void *@var{data}, int @var{size})
1405 To add a block of initialized space, use @code{obstack_grow}, which is
1406 the growing-object analogue of @code{obstack_copy}. It adds @var{size}
1407 bytes of data to the growing object, copying the contents from
1413 @deftypefun void obstack_grow0 (struct obstack *@var{obstack-ptr}, void *@var{data}, int @var{size})
1414 This is the growing-object analogue of @code{obstack_copy0}. It adds
1415 @var{size} bytes copied from @var{data}, followed by an additional null
1421 @deftypefun void obstack_1grow (struct obstack *@var{obstack-ptr}, char @var{c})
1422 To add one character at a time, use the function @code{obstack_1grow}.
1423 It adds a single byte containing @var{c} to the growing object.
1428 @deftypefun void obstack_ptr_grow (struct obstack *@var{obstack-ptr}, void *@var{data})
1429 Adding the value of a pointer one can use the function
1430 @code{obstack_ptr_grow}. It adds @code{sizeof (void *)} bytes
1431 containing the value of @var{data}.
1436 @deftypefun void obstack_int_grow (struct obstack *@var{obstack-ptr}, int @var{data})
1437 A single value of type @code{int} can be added by using the
1438 @code{obstack_int_grow} function. It adds @code{sizeof (int)} bytes to
1439 the growing object and initializes them with the value of @var{data}.
1444 @deftypefun {void *} obstack_finish (struct obstack *@var{obstack-ptr})
1445 When you are finished growing the object, use the function
1446 @code{obstack_finish} to close it off and return its final address.
1448 Once you have finished the object, the obstack is available for ordinary
1449 allocation or for growing another object.
1451 This function can return a null pointer under the same conditions as
1452 @code{obstack_alloc} (@pxref{Allocation in an Obstack}).
1455 When you build an object by growing it, you will probably need to know
1456 afterward how long it became. You need not keep track of this as you grow
1457 the object, because you can find out the length from the obstack just
1458 before finishing the object with the function @code{obstack_object_size},
1459 declared as follows:
1463 @deftypefun int obstack_object_size (struct obstack *@var{obstack-ptr})
1464 This function returns the current size of the growing object, in bytes.
1465 Remember to call this function @emph{before} finishing the object.
1466 After it is finished, @code{obstack_object_size} will return zero.
1469 If you have started growing an object and wish to cancel it, you should
1470 finish it and then free it, like this:
1473 obstack_free (obstack_ptr, obstack_finish (obstack_ptr));
1477 This has no effect if no object was growing.
1479 @cindex shrinking objects
1480 You can use @code{obstack_blank} with a negative size argument to make
1481 the current object smaller. Just don't try to shrink it beyond zero
1482 length---there's no telling what will happen if you do that.
1484 @node Extra Fast Growing
1485 @subsection Extra Fast Growing Objects
1486 @cindex efficiency and obstacks
1488 The usual functions for growing objects incur overhead for checking
1489 whether there is room for the new growth in the current chunk. If you
1490 are frequently constructing objects in small steps of growth, this
1491 overhead can be significant.
1493 You can reduce the overhead by using special ``fast growth''
1494 functions that grow the object without checking. In order to have a
1495 robust program, you must do the checking yourself. If you do this checking
1496 in the simplest way each time you are about to add data to the object, you
1497 have not saved anything, because that is what the ordinary growth
1498 functions do. But if you can arrange to check less often, or check
1499 more efficiently, then you make the program faster.
1501 The function @code{obstack_room} returns the amount of room available
1502 in the current chunk. It is declared as follows:
1506 @deftypefun int obstack_room (struct obstack *@var{obstack-ptr})
1507 This returns the number of bytes that can be added safely to the current
1508 growing object (or to an object about to be started) in obstack
1509 @var{obstack} using the fast growth functions.
1512 While you know there is room, you can use these fast growth functions
1513 for adding data to a growing object:
1517 @deftypefun void obstack_1grow_fast (struct obstack *@var{obstack-ptr}, char @var{c})
1518 The function @code{obstack_1grow_fast} adds one byte containing the
1519 character @var{c} to the growing object in obstack @var{obstack-ptr}.
1524 @deftypefun void obstack_ptr_grow_fast (struct obstack *@var{obstack-ptr}, void *@var{data})
1525 The function @code{obstack_ptr_grow_fast} adds @code{sizeof (void *)}
1526 bytes containing the value of @var{data} to the growing object in
1527 obstack @var{obstack-ptr}.
1532 @deftypefun void obstack_int_grow_fast (struct obstack *@var{obstack-ptr}, int @var{data})
1533 The function @code{obstack_int_grow_fast} adds @code{sizeof (int)} bytes
1534 containing the value of @var{data} to the growing object in obstack
1540 @deftypefun void obstack_blank_fast (struct obstack *@var{obstack-ptr}, int @var{size})
1541 The function @code{obstack_blank_fast} adds @var{size} bytes to the
1542 growing object in obstack @var{obstack-ptr} without initializing them.
1545 When you check for space using @code{obstack_room} and there is not
1546 enough room for what you want to add, the fast growth functions
1547 are not safe. In this case, simply use the corresponding ordinary
1548 growth function instead. Very soon this will copy the object to a
1549 new chunk; then there will be lots of room available again.
1551 So, each time you use an ordinary growth function, check afterward for
1552 sufficient space using @code{obstack_room}. Once the object is copied
1553 to a new chunk, there will be plenty of space again, so the program will
1554 start using the fast growth functions again.
1561 add_string (struct obstack *obstack, const char *ptr, int len)
1565 int room = obstack_room (obstack);
1568 /* @r{Not enough room. Add one character slowly,}
1569 @r{which may copy to a new chunk and make room.} */
1570 obstack_1grow (obstack, *ptr++);
1577 /* @r{Add fast as much as we have room for.} */
1580 obstack_1grow_fast (obstack, *ptr++);
1587 @node Status of an Obstack
1588 @subsection Status of an Obstack
1589 @cindex obstack status
1590 @cindex status of obstack
1592 Here are functions that provide information on the current status of
1593 allocation in an obstack. You can use them to learn about an object while
1598 @deftypefun {void *} obstack_base (struct obstack *@var{obstack-ptr})
1599 This function returns the tentative address of the beginning of the
1600 currently growing object in @var{obstack-ptr}. If you finish the object
1601 immediately, it will have that address. If you make it larger first, it
1602 may outgrow the current chunk---then its address will change!
1604 If no object is growing, this value says where the next object you
1605 allocate will start (once again assuming it fits in the current
1611 @deftypefun {void *} obstack_next_free (struct obstack *@var{obstack-ptr})
1612 This function returns the address of the first free byte in the current
1613 chunk of obstack @var{obstack-ptr}. This is the end of the currently
1614 growing object. If no object is growing, @code{obstack_next_free}
1615 returns the same value as @code{obstack_base}.
1620 @deftypefun int obstack_object_size (struct obstack *@var{obstack-ptr})
1621 This function returns the size in bytes of the currently growing object.
1622 This is equivalent to
1625 obstack_next_free (@var{obstack-ptr}) - obstack_base (@var{obstack-ptr})
1629 @node Obstacks Data Alignment
1630 @subsection Alignment of Data in Obstacks
1631 @cindex alignment (in obstacks)
1633 Each obstack has an @dfn{alignment boundary}; each object allocated in
1634 the obstack automatically starts on an address that is a multiple of the
1635 specified boundary. By default, this boundary is 4 bytes.
1637 To access an obstack's alignment boundary, use the macro
1638 @code{obstack_alignment_mask}, whose function prototype looks like
1643 @deftypefn Macro int obstack_alignment_mask (struct obstack *@var{obstack-ptr})
1644 The value is a bit mask; a bit that is 1 indicates that the corresponding
1645 bit in the address of an object should be 0. The mask value should be one
1646 less than a power of 2; the effect is that all object addresses are
1647 multiples of that power of 2. The default value of the mask is 3, so that
1648 addresses are multiples of 4. A mask value of 0 means an object can start
1649 on any multiple of 1 (that is, no alignment is required).
1651 The expansion of the macro @code{obstack_alignment_mask} is an lvalue,
1652 so you can alter the mask by assignment. For example, this statement:
1655 obstack_alignment_mask (obstack_ptr) = 0;
1659 has the effect of turning off alignment processing in the specified obstack.
1662 Note that a change in alignment mask does not take effect until
1663 @emph{after} the next time an object is allocated or finished in the
1664 obstack. If you are not growing an object, you can make the new
1665 alignment mask take effect immediately by calling @code{obstack_finish}.
1666 This will finish a zero-length object and then do proper alignment for
1669 @node Obstack Chunks
1670 @subsection Obstack Chunks
1671 @cindex efficiency of chunks
1674 Obstacks work by allocating space for themselves in large chunks, and
1675 then parceling out space in the chunks to satisfy your requests. Chunks
1676 are normally 4096 bytes long unless you specify a different chunk size.
1677 The chunk size includes 8 bytes of overhead that are not actually used
1678 for storing objects. Regardless of the specified size, longer chunks
1679 will be allocated when necessary for long objects.
1681 The obstack library allocates chunks by calling the function
1682 @code{obstack_chunk_alloc}, which you must define. When a chunk is no
1683 longer needed because you have freed all the objects in it, the obstack
1684 library frees the chunk by calling @code{obstack_chunk_free}, which you
1687 These two must be defined (as macros) or declared (as functions) in each
1688 source file that uses @code{obstack_init} (@pxref{Creating Obstacks}).
1689 Most often they are defined as macros like this:
1692 #define obstack_chunk_alloc malloc
1693 #define obstack_chunk_free free
1696 Note that these are simple macros (no arguments). Macro definitions with
1697 arguments will not work! It is necessary that @code{obstack_chunk_alloc}
1698 or @code{obstack_chunk_free}, alone, expand into a function name if it is
1699 not itself a function name.
1701 If you allocate chunks with @code{malloc}, the chunk size should be a
1702 power of 2. The default chunk size, 4096, was chosen because it is long
1703 enough to satisfy many typical requests on the obstack yet short enough
1704 not to waste too much memory in the portion of the last chunk not yet used.
1708 @deftypefn Macro int obstack_chunk_size (struct obstack *@var{obstack-ptr})
1709 This returns the chunk size of the given obstack.
1712 Since this macro expands to an lvalue, you can specify a new chunk size by
1713 assigning it a new value. Doing so does not affect the chunks already
1714 allocated, but will change the size of chunks allocated for that particular
1715 obstack in the future. It is unlikely to be useful to make the chunk size
1716 smaller, but making it larger might improve efficiency if you are
1717 allocating many objects whose size is comparable to the chunk size. Here
1718 is how to do so cleanly:
1721 if (obstack_chunk_size (obstack_ptr) < @var{new-chunk-size})
1722 obstack_chunk_size (obstack_ptr) = @var{new-chunk-size};
1725 @node Summary of Obstacks
1726 @subsection Summary of Obstack Functions
1728 Here is a summary of all the functions associated with obstacks. Each
1729 takes the address of an obstack (@code{struct obstack *}) as its first
1733 @item void obstack_init (struct obstack *@var{obstack-ptr})
1734 Initialize use of an obstack. @xref{Creating Obstacks}.
1736 @item void *obstack_alloc (struct obstack *@var{obstack-ptr}, int @var{size})
1737 Allocate an object of @var{size} uninitialized bytes.
1738 @xref{Allocation in an Obstack}.
1740 @item void *obstack_copy (struct obstack *@var{obstack-ptr}, void *@var{address}, int @var{size})
1741 Allocate an object of @var{size} bytes, with contents copied from
1742 @var{address}. @xref{Allocation in an Obstack}.
1744 @item void *obstack_copy0 (struct obstack *@var{obstack-ptr}, void *@var{address}, int @var{size})
1745 Allocate an object of @var{size}+1 bytes, with @var{size} of them copied
1746 from @var{address}, followed by a null character at the end.
1747 @xref{Allocation in an Obstack}.
1749 @item void obstack_free (struct obstack *@var{obstack-ptr}, void *@var{object})
1750 Free @var{object} (and everything allocated in the specified obstack
1751 more recently than @var{object}). @xref{Freeing Obstack Objects}.
1753 @item void obstack_blank (struct obstack *@var{obstack-ptr}, int @var{size})
1754 Add @var{size} uninitialized bytes to a growing object.
1755 @xref{Growing Objects}.
1757 @item void obstack_grow (struct obstack *@var{obstack-ptr}, void *@var{address}, int @var{size})
1758 Add @var{size} bytes, copied from @var{address}, to a growing object.
1759 @xref{Growing Objects}.
1761 @item void obstack_grow0 (struct obstack *@var{obstack-ptr}, void *@var{address}, int @var{size})
1762 Add @var{size} bytes, copied from @var{address}, to a growing object,
1763 and then add another byte containing a null character. @xref{Growing
1766 @item void obstack_1grow (struct obstack *@var{obstack-ptr}, char @var{data-char})
1767 Add one byte containing @var{data-char} to a growing object.
1768 @xref{Growing Objects}.
1770 @item void *obstack_finish (struct obstack *@var{obstack-ptr})
1771 Finalize the object that is growing and return its permanent address.
1772 @xref{Growing Objects}.
1774 @item int obstack_object_size (struct obstack *@var{obstack-ptr})
1775 Get the current size of the currently growing object. @xref{Growing
1778 @item void obstack_blank_fast (struct obstack *@var{obstack-ptr}, int @var{size})
1779 Add @var{size} uninitialized bytes to a growing object without checking
1780 that there is enough room. @xref{Extra Fast Growing}.
1782 @item void obstack_1grow_fast (struct obstack *@var{obstack-ptr}, char @var{data-char})
1783 Add one byte containing @var{data-char} to a growing object without
1784 checking that there is enough room. @xref{Extra Fast Growing}.
1786 @item int obstack_room (struct obstack *@var{obstack-ptr})
1787 Get the amount of room now available for growing the current object.
1788 @xref{Extra Fast Growing}.
1790 @item int obstack_alignment_mask (struct obstack *@var{obstack-ptr})
1791 The mask used for aligning the beginning of an object. This is an
1792 lvalue. @xref{Obstacks Data Alignment}.
1794 @item int obstack_chunk_size (struct obstack *@var{obstack-ptr})
1795 The size for allocating chunks. This is an lvalue. @xref{Obstack Chunks}.
1797 @item void *obstack_base (struct obstack *@var{obstack-ptr})
1798 Tentative starting address of the currently growing object.
1799 @xref{Status of an Obstack}.
1801 @item void *obstack_next_free (struct obstack *@var{obstack-ptr})
1802 Address just after the end of the currently growing object.
1803 @xref{Status of an Obstack}.
1806 @node Variable Size Automatic
1807 @section Automatic Storage with Variable Size
1808 @cindex automatic freeing
1809 @cindex @code{alloca} function
1810 @cindex automatic storage with variable size
1812 The function @code{alloca} supports a kind of half-dynamic allocation in
1813 which blocks are allocated dynamically but freed automatically.
1815 Allocating a block with @code{alloca} is an explicit action; you can
1816 allocate as many blocks as you wish, and compute the size at run time. But
1817 all the blocks are freed when you exit the function that @code{alloca} was
1818 called from, just as if they were automatic variables declared in that
1819 function. There is no way to free the space explicitly.
1821 The prototype for @code{alloca} is in @file{stdlib.h}. This function is
1827 @deftypefun {void *} alloca (size_t @var{size});
1828 The return value of @code{alloca} is the address of a block of @var{size}
1829 bytes of storage, allocated in the stack frame of the calling function.
1832 Do not use @code{alloca} inside the arguments of a function call---you
1833 will get unpredictable results, because the stack space for the
1834 @code{alloca} would appear on the stack in the middle of the space for
1835 the function arguments. An example of what to avoid is @code{foo (x,
1837 @c This might get fixed in future versions of GCC, but that won't make
1838 @c it safe with compilers generally.
1841 * Alloca Example:: Example of using @code{alloca}.
1842 * Advantages of Alloca:: Reasons to use @code{alloca}.
1843 * Disadvantages of Alloca:: Reasons to avoid @code{alloca}.
1844 * GNU C Variable-Size Arrays:: Only in GNU C, here is an alternative
1845 method of allocating dynamically and
1846 freeing automatically.
1849 @node Alloca Example
1850 @subsection @code{alloca} Example
1852 As an example of use of @code{alloca}, here is a function that opens a file
1853 name made from concatenating two argument strings, and returns a file
1854 descriptor or minus one signifying failure:
1858 open2 (char *str1, char *str2, int flags, int mode)
1860 char *name = (char *) alloca (strlen (str1) + strlen (str2) + 1);
1861 stpcpy (stpcpy (name, str1), str2);
1862 return open (name, flags, mode);
1867 Here is how you would get the same results with @code{malloc} and
1872 open2 (char *str1, char *str2, int flags, int mode)
1874 char *name = (char *) malloc (strlen (str1) + strlen (str2) + 1);
1877 fatal ("virtual memory exceeded");
1878 stpcpy (stpcpy (name, str1), str2);
1879 desc = open (name, flags, mode);
1885 As you can see, it is simpler with @code{alloca}. But @code{alloca} has
1886 other, more important advantages, and some disadvantages.
1888 @node Advantages of Alloca
1889 @subsection Advantages of @code{alloca}
1891 Here are the reasons why @code{alloca} may be preferable to @code{malloc}:
1895 Using @code{alloca} wastes very little space and is very fast. (It is
1896 open-coded by the GNU C compiler.)
1899 Since @code{alloca} does not have separate pools for different sizes of
1900 block, space used for any size block can be reused for any other size.
1901 @code{alloca} does not cause storage fragmentation.
1905 Nonlocal exits done with @code{longjmp} (@pxref{Non-Local Exits})
1906 automatically free the space allocated with @code{alloca} when they exit
1907 through the function that called @code{alloca}. This is the most
1908 important reason to use @code{alloca}.
1910 To illustrate this, suppose you have a function
1911 @code{open_or_report_error} which returns a descriptor, like
1912 @code{open}, if it succeeds, but does not return to its caller if it
1913 fails. If the file cannot be opened, it prints an error message and
1914 jumps out to the command level of your program using @code{longjmp}.
1915 Let's change @code{open2} (@pxref{Alloca Example}) to use this
1920 open2 (char *str1, char *str2, int flags, int mode)
1922 char *name = (char *) alloca (strlen (str1) + strlen (str2) + 1);
1923 stpcpy (stpcpy (name, str1), str2);
1924 return open_or_report_error (name, flags, mode);
1929 Because of the way @code{alloca} works, the storage it allocates is
1930 freed even when an error occurs, with no special effort required.
1932 By contrast, the previous definition of @code{open2} (which uses
1933 @code{malloc} and @code{free}) would develop a storage leak if it were
1934 changed in this way. Even if you are willing to make more changes to
1935 fix it, there is no easy way to do so.
1938 @node Disadvantages of Alloca
1939 @subsection Disadvantages of @code{alloca}
1941 @cindex @code{alloca} disadvantages
1942 @cindex disadvantages of @code{alloca}
1943 These are the disadvantages of @code{alloca} in comparison with
1948 If you try to allocate more storage than the machine can provide, you
1949 don't get a clean error message. Instead you get a fatal signal like
1950 the one you would get from an infinite recursion; probably a
1951 segmentation violation (@pxref{Program Error Signals}).
1954 Some non-GNU systems fail to support @code{alloca}, so it is less
1955 portable. However, a slower emulation of @code{alloca} written in C
1956 is available for use on systems with this deficiency.
1959 @node GNU C Variable-Size Arrays
1960 @subsection GNU C Variable-Size Arrays
1961 @cindex variable-sized arrays
1963 In GNU C, you can replace most uses of @code{alloca} with an array of
1964 variable size. Here is how @code{open2} would look then:
1967 int open2 (char *str1, char *str2, int flags, int mode)
1969 char name[strlen (str1) + strlen (str2) + 1];
1970 stpcpy (stpcpy (name, str1), str2);
1971 return open (name, flags, mode);
1975 But @code{alloca} is not always equivalent to a variable-sized array, for
1980 A variable size array's space is freed at the end of the scope of the
1981 name of the array. The space allocated with @code{alloca}
1982 remains until the end of the function.
1985 It is possible to use @code{alloca} within a loop, allocating an
1986 additional block on each iteration. This is impossible with
1987 variable-sized arrays.
1990 @strong{Note:} If you mix use of @code{alloca} and variable-sized arrays
1991 within one function, exiting a scope in which a variable-sized array was
1992 declared frees all blocks allocated with @code{alloca} during the
1993 execution of that scope.
1996 @node Relocating Allocator
1997 @section Relocating Allocator
1999 @cindex relocating memory allocator
2000 Any system of dynamic memory allocation has overhead: the amount of
2001 space it uses is more than the amount the program asks for. The
2002 @dfn{relocating memory allocator} achieves very low overhead by moving
2003 blocks in memory as necessary, on its own initiative.
2006 * Relocator Concepts:: How to understand relocating allocation.
2007 * Using Relocator:: Functions for relocating allocation.
2010 @node Relocator Concepts
2011 @subsection Concepts of Relocating Allocation
2014 The @dfn{relocating memory allocator} achieves very low overhead by
2015 moving blocks in memory as necessary, on its own initiative.
2018 When you allocate a block with @code{malloc}, the address of the block
2019 never changes unless you use @code{realloc} to change its size. Thus,
2020 you can safely store the address in various places, temporarily or
2021 permanently, as you like. This is not safe when you use the relocating
2022 memory allocator, because any and all relocatable blocks can move
2023 whenever you allocate memory in any fashion. Even calling @code{malloc}
2024 or @code{realloc} can move the relocatable blocks.
2027 For each relocatable block, you must make a @dfn{handle}---a pointer
2028 object in memory, designated to store the address of that block. The
2029 relocating allocator knows where each block's handle is, and updates the
2030 address stored there whenever it moves the block, so that the handle
2031 always points to the block. Each time you access the contents of the
2032 block, you should fetch its address anew from the handle.
2034 To call any of the relocating allocator functions from a signal handler
2035 is almost certainly incorrect, because the signal could happen at any
2036 time and relocate all the blocks. The only way to make this safe is to
2037 block the signal around any access to the contents of any relocatable
2038 block---not a convenient mode of operation. @xref{Nonreentrancy}.
2040 @node Using Relocator
2041 @subsection Allocating and Freeing Relocatable Blocks
2044 In the descriptions below, @var{handleptr} designates the address of the
2045 handle. All the functions are declared in @file{malloc.h}; all are GNU
2050 @deftypefun {void *} r_alloc (void **@var{handleptr}, size_t @var{size})
2051 This function allocates a relocatable block of size @var{size}. It
2052 stores the block's address in @code{*@var{handleptr}} and returns
2053 a non-null pointer to indicate success.
2055 If @code{r_alloc} can't get the space needed, it stores a null pointer
2056 in @code{*@var{handleptr}}, and returns a null pointer.
2061 @deftypefun void r_alloc_free (void **@var{handleptr})
2062 This function is the way to free a relocatable block. It frees the
2063 block that @code{*@var{handleptr}} points to, and stores a null pointer
2064 in @code{*@var{handleptr}} to show it doesn't point to an allocated
2070 @deftypefun {void *} r_re_alloc (void **@var{handleptr}, size_t @var{size})
2071 The function @code{r_re_alloc} adjusts the size of the block that
2072 @code{*@var{handleptr}} points to, making it @var{size} bytes long. It
2073 stores the address of the resized block in @code{*@var{handleptr}} and
2074 returns a non-null pointer to indicate success.
2076 If enough memory is not available, this function returns a null pointer
2077 and does not modify @code{*@var{handleptr}}.
2081 @comment No longer available...
2083 @comment @node Memory Warnings
2084 @comment @section Memory Usage Warnings
2085 @comment @cindex memory usage warnings
2086 @comment @cindex warnings of memory almost full
2089 You can ask for warnings as the program approaches running out of memory
2090 space, by calling @code{memory_warnings}. This tells @code{malloc} to
2091 check memory usage every time it asks for more memory from the operating
2092 system. This is a GNU extension declared in @file{malloc.h}.
2096 @comment @deftypefun void memory_warnings (void *@var{start}, void (*@var{warn-func}) (const char *))
2097 Call this function to request warnings for nearing exhaustion of virtual
2100 The argument @var{start} says where data space begins, in memory. The
2101 allocator compares this against the last address used and against the
2102 limit of data space, to determine the fraction of available memory in
2103 use. If you supply zero for @var{start}, then a default value is used
2104 which is right in most circumstances.
2106 For @var{warn-func}, supply a function that @code{malloc} can call to
2107 warn you. It is called with a string (a warning message) as argument.
2108 Normally it ought to display the string for the user to read.
2111 The warnings come when memory becomes 75% full, when it becomes 85%
2112 full, and when it becomes 95% full. Above 95% you get another warning
2113 each time memory usage increases.