Other AliasGC_malloc, GC_malloc_atomic, GC_realloc, GC_enable_incremental, GC_register_finalizer, GC_malloc_ignore_off_page, GC_malloc_atomic_ignore_off_page, GC_set_warn_proc
void * GC_malloc(size_t size);
void GC_free(void *ptr);
void * GC_realloc(void *ptr, size_t size);
+ cc ... -I/usr/include/gc -lgc
DESCRIPTIONGC_malloc and GC_free are plug-in replacements for standard malloc and free. However, GC_malloc will attempt to reclaim inaccessible space automatically by invoking a conservative garbage collector at appropriate points. The collector traverses all data structures accessible by following pointers from the machines registers, stack(s), data, and bss segments. Inaccessible structures will be reclaimed. A machine word is considered to be a valid pointer if it is an address inside an object allocated by GC_malloc or friends.
In most cases it is preferable to call the macros GC_MALLOC, GC_FREE, etc. instead of calling GC_malloc and friends directly. This allows debugging versions of the routines to be substituted by defining GC_DEBUG before including gc.h.
See the documentation in the include file gc_cpp.h for an alternate, C++ specific interface to the garbage collector.
Unlike the standard implementations of malloc, GC_malloc clears the newly allocated storage. GC_malloc_atomic does not. Furthermore, it informs the collector that the resulting object will never contain any pointers, and should therefore not be scanned by the collector.
GC_free can be used to deallocate objects, but its use is optional, and generally discouraged. GC_realloc has the standard realloc semantics. It preserves pointer-free-ness. GC_register_finalizer allows for registration of functions that are invoked when an object becomes inaccessible.
The garbage collector tries to avoid allocating memory at locations that already appear to be referenced before allocation. (Such apparent ``pointers'' are usually large integers and the like that just happen to look like an address.) This may make it hard to allocate very large objects. An attempt to do so may generate a warning.
GC_malloc_ignore_off_page and GC_malloc_atomic_ignore_off_page inform the collector that the client code will always maintain a pointer to near the beginning of the object (within the first 512 bytes), and that pointers beyond that can be ignored by the collector. This makes it much easier for the collector to place large objects. These are recommended for large object allocation. (Objects expected to be larger than about 100KBytes should be allocated this way.)
It is also possible to use the collector to find storage leaks in programs destined to be run with standard malloc/free. The collector can be compiled for thread-safe operation. Unlike standard malloc, it is safe to call malloc after a previous malloc call was interrupted by a signal, provided the original malloc call is not resumed.
The collector may, on rare occasion produce warning messages. On UNIX machines these appear on stderr. Warning messages can be filtered, redirected, or ignored with GC_set_warn_proc This is recommended for production code. See gc.h for details.
Fully portable code should call GC_INIT from the main program before making any other GC calls. On most platforms this does nothing and the collector is initialized on first use. On a few platforms explicit initialization is necessary. And it can never hurt.
Debugging versions of many of the above routines are provided as macros. Their names are identical to the above, but consist of all capital letters. If GC_DEBUG is defined before gc.h is included, these routines do additional checking, and allow the leak detecting version of the collector to produce slightly more useful output. Without GC_DEBUG defined, they behave exactly like the lower-case versions.
On some machines, collection will be performed incrementally after a call to GC_enable_incremental. This may temporarily write protect pages in the heap. See the README file for more information on how this interacts with system calls that write to the heap.
Other facilities not discussed here include limited facilities to support incremental collection on machines without appropriate VM support, provisions for providing more explicit object layout information to the garbage collector, more direct support for ``weak'' pointers, support for ``abortable'' garbage collections during idle time, etc.
AUTHORHans-J. Boehm ([email protected]). Some of the code was written by others, most notably Alan Demers.