Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.c > #6872 > unrolled thread

Should you free your pointers?

Back to article view | Back to comp.lang.c

Contents

  Should you free your pointers? Billy Mays <noway@nohow.com> - 2011-06-24 16:41 -0400
    Re: Should you free your pointers? Shao Miller <sha0.miller@gmail.com> - 2011-06-24 16:51 -0400
      Re: Should you free your pointers? Shao Miller <sha0.miller@gmail.com> - 2011-06-24 16:56 -0400
      Re: Should you free your pointers? falk@rahul.net (Edward A. Falk) - 2011-06-26 04:09 +0000
    Re: Should you free your pointers? Nobody <nobody@nowhere.com> - 2011-06-24 21:53 +0100
      Re: Should you free your pointers? Martin Ambuhl <mambuhl@earthlink.net> - 2011-06-24 19:06 -0400
    Re: Should you free your pointers? Ben Pfaff <blp@cs.stanford.edu> - 2011-06-24 14:05 -0700
      Re: Should you free your pointers? David Mathog <dmathog@gmail.com> - 2011-06-28 09:36 -0700
    Re: Should you free your pointers? Ian Collins <ian-news@hotmail.com> - 2011-06-25 09:31 +1200
    Re: Should you free your pointers? Martin Ambuhl <mambuhl@earthlink.net> - 2011-06-24 19:04 -0400
    Re: Should you free your pointers? Eric Sosman <esosman@ieee-dot-org.invalid> - 2011-06-24 21:19 -0400
      Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-25 15:15 -0700
        Re: Should you free your pointers? Keith Thompson <kst-u@mib.org> - 2011-06-25 15:44 -0700
          Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-26 12:56 -0700
            Re: Should you free your pointers? Keith Thompson <kst-u@mib.org> - 2011-06-26 13:08 -0700
              Re: Should you free your pointers? Geoff <geoff@invalid.invalid> - 2011-06-26 14:33 -0700
              Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-26 16:27 -0700
          Re: Should you free your pointers? James Kuyper <jameskuyper@verizon.net> - 2011-06-26 19:39 -0400
            Re: Should you free your pointers? Keith Thompson <kst-u@mib.org> - 2011-06-26 19:28 -0700
              Re: Should you free your pointers? James Waldby <not@valid.invalid> - 2011-06-27 05:42 +0000
              Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-27 12:59 -0700
                Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-27 13:06 -0700
                Re: Should you free your pointers? Seebs <usenet-nospam@seebs.net> - 2011-06-27 20:26 +0000
                Re: Should you free your pointers? Joe Pfeiffer <pfeiffer@cs.nmsu.edu> - 2011-06-27 14:28 -0600
                Re: Should you free your pointers? pacman@kosh.dhis.org (Alan Curry) - 2011-06-27 23:29 +0000
                Re: Should you free your pointers? pete <pfiland@mindspring.com> - 2011-06-27 21:08 -0400
                Re: Should you free your pointers? "J. J. Farrell" <jjf@bcs.org.uk> - 2011-06-28 04:05 +0100
                  Re: Should you free your pointers? Billy Mays <noway@nohow.com> - 2011-06-28 00:25 -0400
                    Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-28 01:23 -0700
                  Re: Should you free your pointers? Shao Miller <sha0.miller@gmail.com> - 2011-06-28 01:00 -0500
                  Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-28 01:22 -0700
                    Re: Should you free your pointers? Joe Pfeiffer <pfeiffer@cs.nmsu.edu> - 2011-06-28 10:02 -0600
                    Re: Should you free your pointers? "J. J. Farrell" <jjf@bcs.org.uk> - 2011-06-29 02:41 +0100
                      Re: Should you free your pointers? Joe Pfeiffer <pfeiffer@cs.nmsu.edu> - 2011-06-28 19:47 -0600
                Re: Should you free your pointers? James Kuyper <jameskuyper@verizon.net> - 2011-06-28 08:21 -0400
                  Re: Should you free your pointers? Michael Press <rubrum@pacbell.net> - 2011-06-30 15:43 -0700
                Re: Should you free your pointers? Keith Thompson <kst-u@mib.org> - 2011-06-28 22:36 -0700
    Re: Should you free your pointers? James Kuyper <jameskuyper@verizon.net> - 2011-06-24 22:07 -0400
      Re: Should you free your pointers? Nobody <nobody@nowhere.com> - 2011-06-25 15:27 +0100
    Re: Should you free your pointers? China Blue Dolls <chine.bleu@yahoo.com> - 2011-06-27 18:16 -0700
    Re: Should you free your pointers? DDD <1983ddd@gmail.com> - 2011-06-28 19:41 -0700

Page 1 of 3  [1] 2 3  Next page →

#6872 — Should you free your pointers?

In the case where you know that your program will be terminating after 
it is done using the pointers, should you even bother freeing the space up?

Example:
int main(void)
{
     int i, * data;
     data = malloc(sizeof(int) * 100);
     if(data) {
         for(i = 0; i < 100; i++) {
             data[i] = i'
         }
     }
     /* free(data) ? */
     return 0;
}

--
Bill

[toc] | [next] | [standalone]

#6877

On 6/24/2011 16:41, Billy Mays wrote:
> In the case where you know that your program will be terminating after
> it is done using the pointers, should you even bother freeing the space up?
>
> Example:
> int main(void)
> {
>     int i, * data;
>     data = malloc(sizeof(int) * 100);
>     if(data) {
>         for(i = 0; i < 100; i++) {
>             data[i] = i'
>         }
>     }
>     /* free(data) ? */
>     return 0;
> }

I've seen arguments for:
- Program performance
- Program size

And against:
- Programmer habit
- Cases where program termination doesn't free memory allocated by the 
program

I think it's a worth-while habit to 'free' them, in general.

[toc] | [prev] | [next] | [standalone]

#6879

On 6/24/2011 16:51, Shao Miller wrote:
> On 6/24/2011 16:41, Billy Mays wrote:
>> In the case where you know that your program will be terminating after
>> it is done using the pointers, should you even bother freeing the
>> space up?
>>
>> Example:
>> int main(void)
>> {
>>     int i, * data;
>>     data = malloc(sizeof(int) * 100);
>>     if(data) {
>>         for(i = 0; i < 100; i++) {
>>             data[i] = i'
>>         }
>>     }
>>     /* free(data) ? */
>>     return 0;
>> }
>
> I've seen arguments for:
> - Program performance
> - Program size
>
> And against:
> - Programmer habit
> - Cases where program termination doesn't free memory allocated by the
> program
>
> I think it's a worth-while habit to 'free' them, in general.

Uh, I meant "for not bothering" and "against not bothering." :)

[toc] | [prev] | [next] | [standalone]

#6951

In article <iu2tf6$rm7$1@dont-email.me>,
Shao Miller  <sha0.miller@gmail.com> wrote:
>On 6/24/2011 16:41, Billy Mays wrote:
>> In the case where you know that your program will be terminating after
>> it is done using the pointers, should you even bother freeing the space up?
>
>And against:
>- Programmer habit
>- Cases where program termination doesn't free memory allocated by the 
>program

- Chances that your code will be re-purposed into some larger
program that *won't* terminate any time soon and which will call
your code repeatedly.

-- 
	-Ed Falk, falk@despams.r.us.com
	http://thespamdiaries.blogspot.com/

[toc] | [prev] | [next] | [standalone]

#6878

On Fri, 24 Jun 2011 16:41:00 -0400, Billy Mays wrote:

> In the case where you know that your program will be terminating after 
> it is done using the pointers, should you even bother freeing the space up?

No.

In complex programs where freeing data may be mixed with other clean-up,
I may set __free_hook to an empty function when termination is imminent.

[toc] | [prev] | [next] | [standalone]

#6908

On 6/24/2011 4:53 PM, Nobody wrote:
> On Fri, 24 Jun 2011 16:41:00 -0400, Billy Mays wrote:
>
>> In the case where you know that your program will be terminating after
>> it is done using the pointers, should you even bother freeing the space up?
>
> No.
>
> In complex programs where freeing data may be mixed with other clean-up,
> I may set __free_hook to an empty function when termination is imminent.
>

It is appropriate that this incorrect answer came from "Nobody".

[toc] | [prev] | [next] | [standalone]

#6882

Billy Mays <noway@nohow.com> writes:

> In the case where you know that your program will be terminating after
> it is done using the pointers, should you even bother freeing the
> space up?

One reason that I sometimes bother is that it's very clear that a
program that frees all of its memory does not have a memory leak
(when I use a tool that reports memory statistics at program
exit).
-- 
Ben Pfaff 
http://benpfaff.org

[toc] | [prev] | [next] | [standalone]

#7154

On Jun 24, 2:05 pm, Ben Pfaff <b...@cs.stanford.edu> wrote:

> One reason that I sometimes bother is that it's very clear that a
> program that frees all of its memory does not have a memory leak
> (when I use a tool that reports memory statistics at program
> exit).

Exactly.  But it isn't hard to have it both ways.  In one of my
programs during development it looks like:

/* For memory leak testing
*/
free()
free()
...

and in production

/* For memory leak testing
free()
free()
...
*/

This could (maybe should) be done with an ifdef too.  Anyway, this
particular program has complex memory structures, so it is easy to
introduce memory leaks and access violations, and the way I find these
is by running a large test suite using valgrind.  (As opposed to the
same large testsuite without valgrind for routine testing.)

As an aside, finding memory leaks in tests that are forced to exit
through:

/* awful unrecoverable problem, exit immediately */
exit(EXIT_FAILURE);

is problematical, since at the exit point the code will rarely have
access to all of the data structures that (might) need to be freed.
The end result
is that in those sections of the testsuite under valgrind it looks
like a memory leak whether or not there really is one.

Regards,

David Mathog

[toc] | [prev] | [next] | [standalone]

#6889

On 06/25/11 08:41 AM, Billy Mays wrote:
> In the case where you know that your program will be terminating after
> it is done using the pointers, should you even bother freeing the space up?

Code fragments that will never be reused often are, so get into good 
habits early!

-- 
Ian Collins

[toc] | [prev] | [next] | [standalone]

#6907

On 6/24/2011 4:41 PM, Billy Mays wrote:
> In the case where you know that your program will be terminating after
> it is done using the pointers, should you even bother freeing the space up?

You should always free any space you have allocated.  There is no 
guarantee that the OS will do it for you.  Even if the OS you are 
currently using claims it will do so, relying on such behavior makes 
your code non-portable and there is no warrant in the C standard for 
expecting otherwise (since the C standard can hardly decree such this 
for the OS).

> Example:

No example is needed.

[toc] | [prev] | [next] | [standalone]

#6911

On 6/24/2011 4:41 PM, Billy Mays wrote:
> In the case where you know that your program will be terminating after
> it is done using the pointers, should you even bother freeing the space up?

     This is Question 7.24 on the comp.lang.c Frequently Asked
Questions (FAQ) page at <http://www.c-faq.com/>.

-- 
Eric Sosman
esosman@ieee-dot-org.invalid

[toc] | [prev] | [next] | [standalone]

#6942

In article <iu3d2t$qlh$1@dont-email.me>,
 Eric Sosman <esosman@ieee-dot-org.invalid> wrote:

> On 6/24/2011 4:41 PM, Billy Mays wrote:
> > In the case where you know that your program will be terminating after
> > it is done using the pointers, should you even bother freeing the space up?
> 
>      This is Question 7.24 on the comp.lang.c Frequently Asked
> Questions (FAQ) page at <http://www.c-faq.com/>.

FAQ was once frequently answered questions.

-- 
Michael Press

[toc] | [prev] | [next] | [standalone]

#6943

Michael Press <rubrum@pacbell.net> writes:
> In article <iu3d2t$qlh$1@dont-email.me>,
>  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
>
>> On 6/24/2011 4:41 PM, Billy Mays wrote:
>> > In the case where you know that your program will be terminating
>> > after it is done using the pointers, should you even bother freeing
>> > the space up?
>> 
>>      This is Question 7.24 on the comp.lang.c Frequently Asked
>> Questions (FAQ) page at <http://www.c-faq.com/>.
>
> FAQ was once frequently answered questions.

Um, it still is.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
Nokia
"We must do something.  This is something.  Therefore, we must do this."
    -- Antony Jay and Jonathan Lynn, "Yes Minister"

[toc] | [prev] | [next] | [standalone]

#6983

In article <lnsjqxbkfi.fsf@nuthaus.mib.org>,
 Keith Thompson <kst-u@mib.org> wrote:

> Michael Press <rubrum@pacbell.net> writes:
> > In article <iu3d2t$qlh$1@dont-email.me>,
> >  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
> >
> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
> >> > In the case where you know that your program will be terminating
> >> > after it is done using the pointers, should you even bother freeing
> >> > the space up?
> >> 
> >>      This is Question 7.24 on the comp.lang.c Frequently Asked
> >> Questions (FAQ) page at <http://www.c-faq.com/>.
> >
> > FAQ was once frequently answered questions.
> 
> Um, it still is.

Um, that is what I am implying.

-- 
Michael Press

[toc] | [prev] | [next] | [standalone]

#6984

Michael Press <rubrum@pacbell.net> writes:
> In article <lnsjqxbkfi.fsf@nuthaus.mib.org>,
>  Keith Thompson <kst-u@mib.org> wrote:
>
>> Michael Press <rubrum@pacbell.net> writes:
>> > In article <iu3d2t$qlh$1@dont-email.me>,
>> >  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
>> >
>> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
>> >> > In the case where you know that your program will be terminating
>> >> > after it is done using the pointers, should you even bother freeing
>> >> > the space up?
>> >> 
>> >>      This is Question 7.24 on the comp.lang.c Frequently Asked
>> >> Questions (FAQ) page at <http://www.c-faq.com/>.
>> >
>> > FAQ was once frequently answered questions.
>> 
>> Um, it still is.
>
> Um, that is what I am implying.

I'm afraid I've missed whatever point you're making.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
Nokia
"We must do something.  This is something.  Therefore, we must do this."
    -- Antony Jay and Jonathan Lynn, "Yes Minister"

[toc] | [prev] | [next] | [standalone]

#6989

On Sun, 26 Jun 2011 13:08:24 -0700, Keith Thompson <kst-u@mib.org>
wrote:

>Michael Press <rubrum@pacbell.net> writes:
>> In article <lnsjqxbkfi.fsf@nuthaus.mib.org>,
>>  Keith Thompson <kst-u@mib.org> wrote:
>>
>>> Michael Press <rubrum@pacbell.net> writes:
>>> > In article <iu3d2t$qlh$1@dont-email.me>,
>>> >  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
>>> >
>>> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
>>> >> > In the case where you know that your program will be terminating
>>> >> > after it is done using the pointers, should you even bother freeing
>>> >> > the space up?
>>> >> 
>>> >>      This is Question 7.24 on the comp.lang.c Frequently Asked
>>> >> Questions (FAQ) page at <http://www.c-faq.com/>.
>>> >
>>> > FAQ was once frequently answered questions.
>>> 
>>> Um, it still is.
>>
>> Um, that is what I am implying.
>
>I'm afraid I've missed whatever point you're making.

I think it means it was once a frequently answered question, now it's
in the FAQ so it no longer needs to be frequently answered. People can
read the FAQ and don't need to ask it in the group. Either that, or he
thinks the FAQ doesn't answer the question but FAQ stands for
Frequently Asked Questions, without answers.

[toc] | [prev] | [next] | [standalone]

#6999

In article <lnei2gbbk7.fsf@nuthaus.mib.org>,
 Keith Thompson <kst-u@mib.org> wrote:

> Michael Press <rubrum@pacbell.net> writes:
> > In article <lnsjqxbkfi.fsf@nuthaus.mib.org>,
> >  Keith Thompson <kst-u@mib.org> wrote:
> >
> >> Michael Press <rubrum@pacbell.net> writes:
> >> > In article <iu3d2t$qlh$1@dont-email.me>,
> >> >  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
> >> >
> >> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
> >> >> > In the case where you know that your program will be terminating
> >> >> > after it is done using the pointers, should you even bother freeing
> >> >> > the space up?
> >> >> 
> >> >>      This is Question 7.24 on the comp.lang.c Frequently Asked
> >> >> Questions (FAQ) page at <http://www.c-faq.com/>.
> >> >
> >> > FAQ was once frequently answered questions.
> >> 
> >> Um, it still is.
> >
> > Um, that is what I am implying.
> 
> I'm afraid I've missed whatever point you're making.

Reminding ES of this matter.

-- 
Michael Press

[toc] | [prev] | [next] | [standalone]

#7002

On 06/25/2011 06:44 PM, Keith Thompson wrote:
> Michael Press <rubrum@pacbell.net> writes:
>> In article <iu3d2t$qlh$1@dont-email.me>,
>>  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
...
>>>      This is Question 7.24 on the comp.lang.c Frequently Asked
>>> Questions (FAQ) page at <http://www.c-faq.com/>.
>>
>> FAQ was once frequently answered questions.
> 
> Um, it still is.

I presume that he's commenting on the difference between "Asked" and
"Answered".
-- 
James Kuyper

[toc] | [prev] | [next] | [standalone]

#7011

James Kuyper <jameskuyper@verizon.net> writes:
> On 06/25/2011 06:44 PM, Keith Thompson wrote:
>> Michael Press <rubrum@pacbell.net> writes:
>>> In article <iu3d2t$qlh$1@dont-email.me>,
>>>  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
> ...
>>>>      This is Question 7.24 on the comp.lang.c Frequently Asked
>>>> Questions (FAQ) page at <http://www.c-faq.com/>.
>>>
>>> FAQ was once frequently answered questions.
>> 
>> Um, it still is.
>
> I presume that he's commenting on the difference between "Asked" and
> "Answered".

I must admit I didn't notice that he changed the word from "asked" to
"answered".

Now that I've noticed, I'm not sure I've ever seen FAQ expanded
to "Frequently Answered Questions" rather than "Frequently Asked
Questions", and I still have no idea what Michael is getting at.
Is he suggesting that Eric should have done something other than
cite the FAQ?

Sorry if I'm being dense.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
Nokia
"We must do something.  This is something.  Therefore, we must do this."
    -- Antony Jay and Jonathan Lynn, "Yes Minister"

[toc] | [prev] | [next] | [standalone]

#7024

On Sun, 26 Jun 2011 19:28:02 -0700, Keith Thompson wrote:
> James Kuyper <jameskuyper@verizon.net> writes:
>> On 06/25/2011 06:44 PM, Keith Thompson wrote:
>>> Michael Press <rubrum@pacbell.net> writes: ...
>>>>  Eric Sosman <esosman@ieee-dot-org.invalid> wrote:
>> ...
>>>>>      This is Question 7.24 on the comp.lang.c Frequently Asked
>>>>> Questions (FAQ) page at <http://www.c-faq.com/>.
>>>>
>>>> FAQ was once frequently answered questions.
>>> 
>>> Um, it still is.
>>
>> I presume that he's commenting on the difference between "Asked" and
>> "Answered".
> 
> I must admit I didn't notice that he changed the word from "asked" to
> "answered".
> 
> Now that I've noticed, I'm not sure I've ever seen FAQ expanded to
> "Frequently Answered Questions" rather than "Frequently Asked
> Questions", and I still have no idea what Michael is getting at. Is he
> suggesting that Eric should have done something other than cite the FAQ?

Michael Press has expanded upon his original statement by the pellucid 
reply, "Um, that is what I am implying" to your comment "Um, it still 
is", and by the eloquent remark, "Reminding ES of this matter" in 
response to your comment, "I'm afraid I've missed whatever point you're 
making".  Given the clarity and eloquence of those replies, how can
you ask more of Press?

> Sorry if I'm being dense.

Ok.

-- 
jiw

[toc] | [prev] | [next] | [standalone]

Page 1 of 3  [1] 2 3  Next page →

Back to top | Article view | comp.lang.c

csiph-web