Article View: vmsnet.internals
Article #361Re: malloc() as a user defined function
From: Dave Schaumann
Date: Sun, 11 Oct 1992 05:47
Date: Sun, 11 Oct 1992 05:47
59 lines
2622 bytes
2622 bytes
In article <Bvv6pH.Gzy@cs.psu.edu>, ytang@red (Yuan-Ling Tang) writes:
> In programming on dynamic memory allocation, I prefer doing check on
> every malloc(). I.e.:
>
> if ((p=(SOME_TYPE *)malloc(sizeof(SOME_TYPE))) == NULL) {
> printf("Malloc error.\n");
> exit(0);
> }
>
> It turned out it's a pain to type in the above four lines if I have
> lots of malloc() invocations in the program, also, the program looks
> ugly. So I tried to use a function called my_malloc():
>
> SOME_TYPE *my_malloc()
[deleted, coded as above]
>
> [...] what if I have many distinct types of memory space to allocate?
Here's an idea I cribbed off of a compiler I worked on last semester (the
source is up for ftp, so it should be OK to use & distribute freely):
Somewhere, I have the following function:
void *mustalloc( size_t size ) {
void *t = malloc(size) ;
if( t == NULL ) FATAL( "malloc returns NULL" ) ;
return t ;
}
FATAL(), of course, is just a simple print 'n' exit routine, which I made
a seperate subroutine since such a thing is often genrally useful in a
large program.
Along with it, I have the following macros:
extern void *mustalloc( size_t ) ;
#define new(t ) ((t *)mustalloc( sizeof(t) ))
#define anew(t,s) ((t *)mustalloc( sizeof(t) * (s) ))
With this, I can write code like
struct foo *p = new(struct FOO) ;
struct foo foo_array = anew(struct FOO, 100) ;
This provides the advantages of a clean interface to a properly typed
dynamic allocation routine, while hiding the ugly details. An added
benefit is that if you do all your allocation through mustalloc(), you
automatically have a place to put malloc() wrapper code if you need to
do malloc()/free() debugging (or a place to do garbage collection, if
you prefer).
I've been using this pretty much since I saw it. IMHO, it works really
great, and I would encourage others to try it too.
--
Caught an internal error--.brainrc restored
Technical Details
Message-ID:
<1992Oct11.054708.26380@organpipe.uug.arizona.edu>Path:
arc1-ca.pugleaf.net!sparky!uunet!gatech!ncar!noao!amethyst!organpipe.uug.arizona.edu!newsReferences:
<Bvv6pH.Gzy@cs.psu.edu>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!gatech!ncar!noao!amethyst!organpipe.uug.arizona.edu!news From: dave@cs.arizona.edu (Dave Schaumann) Newsgroups: vmsnet.internals,vmsnet.misc,comp.lang.c Subject: Re: malloc() as a user defined function Summary: a new() idea Message-ID: <1992Oct11.054708.26380@organpipe.uug.arizona.edu> Date: 11 Oct 92 05:47:08 GMT References: <Bvv6pH.Gzy@cs.psu.edu> Sender: news@organpipe.uug.arizona.edu Reply-To: dave@cs.arizona.edu (Dave Schaumann) Followup-To: vmsnet.internals Organization: University of Arizona Lines: 59 In-Reply-To: ytang@red.crayola.cs.psu.edu (Yuan-Ling Tang)