Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!cs.utexas.edu!uunet!europa.asd.contel.com!sura.net!haven.umd.edu!umd5!jupiter.astro.umd.edu!jjk
From: jjk@jupiter.astro.umd.edu (Jim Klavetter)
Newsgroups: comp.lang.c
Subject: memory management between functions
Summary: how?
Message-ID: <8504@umd5.umd.edu>
Date: 29 Apr 91 07:50:36 GMT
Sender: news@umd5.umd.edu
Reply-To: jjk@astro.umde.edu (Jim Klavetter)
Distribution: na
Organization: U of Md., Astronomy Program
Lines: 67

A few weeks ago I asked about where to allocate memory, in main()
or in the function.  The general idea I got back was that it
doesn't matter as long as you do it correctly (either place is
ok).

I wanted to make the functions completely autonomous, so this is
what I did (in some c-like metacode)

[includes]
double *subtract(a, b, n)
int n;
double *a, *b;
{
int i;
double *c;

[malloc c array here]

for(i=0; i<n; i++)
	c[i]=a[i]-b[i];
free(c);
return(c);
}

(Again, don't worry about efficiency, this isn't the code, just an
example of the malloc() and free() that is important).

The main program would have the fragments:

[malloc result]

result=subtract(vector1, vector2, n);

...

Conceptually, I like the above.  The problem is that it doesn't
work quite as expected:  when I take a look at the pointer result
(that is &*result, not the result of the pointer) before and after
the function it CAN (but not always does) change.  OK, I know I am
freeing the c array, but the very next thing I do is return it, so
I thought the result pointer would have the information
uncorrupted.

The question:  am I on the right track, or is it completely
impossible the way in which I am doing it?  I am calling subtract
(and many other vector operations) many times so I need to worry
about memory management.

In my first posting, someone offered the advice that the call to
subtract should be:

double *subtract(a, b, n, result)

where result is an "empty" array that will be used to return the
subtracted vector.  I think this would solve the problem, but I am
reluctant to use this idea because of the duplication (as well as
the fortranish look).  Is this NECESSARY duplication that I must
live with?

Thanks to all who helped me out enough last time to get into
trouble this time.

jjk@astro.umd.edu also for Athabasca and Reudi
Jim Klavetter
Astronomy
UMD
College Park, MD  20742