Should you free your pointers?
Billy Mays
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
Shao Miller
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.
Nobody
> 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.
Shao Miller
Uh, I meant "for not bothering" and "against not bothering." :)
Ben Pfaff
> 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
Ian Collins
> 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
Martin Ambuhl
> 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.
Martin Ambuhl
It is appropriate that this incorrect answer came from "Nobody".
Eric Sosman
> 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
eso...@ieee-dot-org.invalid
James Kuyper
It's not necessary, but I think it's easier to not make a special case
of main(); the rules for creating well-written code are easier to
remember if they don't have more special cases than they have to.
Also, there's always the possibility that code from main() may someday
get moved off to another function; if that code already contains
properly connected malloc() and free() calls, that's one less thing to
worry about during the move.
--
James Kuyper
Nobody
>> 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?
> It's not necessary, but I think it's easier to not make a special case
> of main(); the rules for creating well-written code are easier to
> remember if they don't have more special cases than they have to.
Sometimes, explicitly freeing memory requires added complexity in order
to keep track of what needs to be freed. E.g. if you have linked data
structures where you could have multiple pointers to a single block of
memory, you have to take steps to avoid multiple free()s of a single
block. Similarly if pointers can be to either static or dynamic data.
Adding such complexity is especially pointless if it only comes into play
upon termination, e.g. if you have complex structures which must exist in
order for the program to perform its normal function.
> Also, there's always the possibility that code from main() may someday
> get moved off to another function; if that code already contains
> properly connected malloc() and free() calls, that's one less thing to
> worry about during the move.
If the structure of the code is such that "pairing" malloc and free makes
sense, I'll do it. But in that case, I may set __free_hook (glibc) or
similar to point to an empty function once termination is assured. This
also deals with external libraries which typically have to free()
everything upon finalisation to be on the safe side.
Meticulously free()ing memory when you know that the program is about to
terminate is like rearranging the deckchairs as the Titanic is sinking.
Such operations can often take far longer than is immediately apparent, as
you often end up walking through memory which has long since been swapped
out.
Michael Press
Eric Sosman <eso...@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
Keith Thompson
Um, it still is.
--
Keith Thompson (The_Other_Keith) ks...@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"
Edward A. Falk
Shao Miller <sha0....@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?
>
>- 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, fa...@despams.r.us.com
http://thespamdiaries.blogspot.com/
Michael Press
Keith Thompson <ks...@mib.org> wrote:
> Michael Press <rub...@pacbell.net> writes:
> > In article <iu3d2t$qlh$1...@dont-email.me>,
> > Eric Sosman <eso...@ieee-dot-org.invalid> wrote:
> >
> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
> >> > In the case where you know that your program will be terminating
> >> > after it is done using the pointers, should you even bother freeing
> >> > the space up?
> >>
> >> This is Question 7.24 on the comp.lang.c Frequently Asked
> >> Questions (FAQ) page at <http://www.c-faq.com/>.
> >
> > FAQ was once frequently answered questions.
>
> Um, it still is.
Um, that is what I am implying.
--
Michael Press
Keith Thompson
> In article <lnsjqxb...@nuthaus.mib.org>,
> Keith Thompson <ks...@mib.org> wrote:
>
>> Michael Press <rub...@pacbell.net> writes:
>> > In article <iu3d2t$qlh$1...@dont-email.me>,
>> > Eric Sosman <eso...@ieee-dot-org.invalid> wrote:
>> >
>> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
>> >> > In the case where you know that your program will be terminating
>> >> > after it is done using the pointers, should you even bother freeing
>> >> > the space up?
>> >>
>> >> This is Question 7.24 on the comp.lang.c Frequently Asked
>> >> Questions (FAQ) page at <http://www.c-faq.com/>.
>> >
>> > FAQ was once frequently answered questions.
>>
>> Um, it still is.
>
> Um, that is what I am implying.
I'm afraid I've missed whatever point you're making.
Geoff
wrote:
>Michael Press <rub...@pacbell.net> writes:
>> In article <lnsjqxb...@nuthaus.mib.org>,
>> Keith Thompson <ks...@mib.org> wrote:
>>
>>> Michael Press <rub...@pacbell.net> writes:
>>> > In article <iu3d2t$qlh$1...@dont-email.me>,
>>> > Eric Sosman <eso...@ieee-dot-org.invalid> wrote:
>>> >
>>> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
>>> >> > In the case where you know that your program will be terminating
>>> >> > after it is done using the pointers, should you even bother freeing
>>> >> > the space up?
>>> >>
>>> >> This is Question 7.24 on the comp.lang.c Frequently Asked
>>> >> Questions (FAQ) page at <http://www.c-faq.com/>.
>>> >
>>> > FAQ was once frequently answered questions.
>>>
>>> Um, it still is.
>>
>> Um, that is what I am implying.
>
>I'm afraid I've missed whatever point you're making.
I think it means it was once a frequently answered question, now it's
in the FAQ so it no longer needs to be frequently answered. People can
read the FAQ and don't need to ask it in the group. Either that, or he
thinks the FAQ doesn't answer the question but FAQ stands for
Frequently Asked Questions, without answers.
Michael Press
Keith Thompson <ks...@mib.org> wrote:
> Michael Press <rub...@pacbell.net> writes:
> > In article <lnsjqxb...@nuthaus.mib.org>,
> > Keith Thompson <ks...@mib.org> wrote:
> >
> >> Michael Press <rub...@pacbell.net> writes:
> >> > In article <iu3d2t$qlh$1...@dont-email.me>,
> >> > Eric Sosman <eso...@ieee-dot-org.invalid> wrote:
> >> >
> >> >> On 6/24/2011 4:41 PM, Billy Mays wrote:
> >> >> > In the case where you know that your program will be terminating
> >> >> > after it is done using the pointers, should you even bother freeing
> >> >> > the space up?
> >> >>
> >> >> This is Question 7.24 on the comp.lang.c Frequently Asked
> >> >> Questions (FAQ) page at <http://www.c-faq.com/>.
> >> >
> >> > FAQ was once frequently answered questions.
> >>
> >> Um, it still is.
> >
> > Um, that is what I am implying.
>
> I'm afraid I've missed whatever point you're making.
Reminding ES of this matter.
--
Michael Press
James Kuyper
> Michael Press <rub...@pacbell.net> writes:
>> In article <iu3d2t$qlh$1...@dont-email.me>,
>> Eric Sosman <eso...@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
Keith Thompson
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.
James Waldby
> James Kuyper <james...@verizon.net> writes:
>> On 06/25/2011 06:44 PM, Keith Thompson wrote:
>>>> Eric Sosman <eso...@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
Michael Press
Keith Thompson <ks...@mib.org> wrote:
> James Kuyper <james...@verizon.net> writes:
> > On 06/25/2011 06:44 PM, Keith Thompson wrote:
> >> Michael Press <rub...@pacbell.net> writes:
> >>> In article <iu3d2t$qlh$1...@dont-email.me>,
> >>> Eric Sosman <eso...@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.
Originally FAQ stood for frequently answered questions.
Most people do not seem to know that, even more than
I thought going in. During the transition, some people
attempted to stem the tide.
--
Michael Press
Michael Press
Michael Press <rub...@pacbell.net> wrote:
A web search for "frequently answered questions"
finds a myriad of sites where the list of frequently
answered questions is titled "Frequently Answered Questions."
--
Michael Press
Seebs
> Originally FAQ stood for frequently answered questions.
Huh! Before my time, I guess.
I would argue that given the existence of a FAQ, most questions on it
are much more frequently asked than answered.
-s
--
Copyright 2011, all wrongs reversed. Peter Seebach / usenet...@seebs.net
http://www.seebs.net/log/ <-- lawsuits, religion, and funny pictures
http://en.wikipedia.org/wiki/Fair_Game_(Scientology) <-- get educated!
I am not speaking for my employer, although they do rent some of my opinions.
Joe Pfeiffer
FWIW, it's been "Asked" for as long as I can remember (and I was on
usenet by 1980). Checking, the jargon file has "Asked".
Alan Curry
>Most people do not seem to know that, even more than
>I thought going in. During the transition, some people
>attempted to stem the tide.
I don't remember that, but I do remember "FAQ" being singular, and a regular
posting containing FAQs and answers being called a "FAQL" (FAQ List). I'm not
sure when that died out.
--
Alan Curry
pete
> Originally FAQ stood for frequently answered questions.
> Most people do not seem to know that, even more than
> I thought going in. During the transition, some people
> attempted to stem the tide.
I am one of those "Most people".
I have always known FAQ to mean Frequently Asked Questions.
--
pete
J. J. Farrell
Evidence please. I've known it to be an abbreviation of "Frequently
Asked Questions" for several decades, this is the first time I've seen a
suggestion that it is "Frequently Answered Questions".
Billy Mays
I think "Asked" fits better since the complement would be Infrequently
Asked Questions rather than Frequently Unanswered Questions.
--
Bill
Shao Miller
>
> Evidence please. I've known it to be an abbreviation of "Frequently
> Asked Questions" for several decades, this is the first time I've seen a
> suggestion that it is "Frequently Answered Questions".
For what it's worth, Mr. J. de Boyne Pollard prefers "Frequently Given
Answers"[1]. I can't say if that's at all relevant.
[1] http://homepage.ntlworld.com/jonathan.deboynepollard/FGA/
Michael Press
The Wikipedia article on FAQ documents this usage.
A web search for "frequently answered questions"
turns up a myriad of sites where the list of frequently
Michael Press
Billy Mays <no...@nohow.com> wrote:
There are frequently asked questions that
never get into the FAQL such as
"Are you serious?"
"What's your point?"
"Is it congenital or were you dropped on your head?"
--
Michael Press
James Kuyper
...
> Originally FAQ stood for frequently answered questions.
> Most people do not seem to know that, even more than
> I thought going in. During the transition, some people
> attempted to stem the tide.
According to Google Groups, the oldest usenet message containing
"Frequently Answered Questions" was dated 1989-03-22. It could only find
14 such messages between then and 1991-01-1.
The oldest message it reports containing the phrase "Frequently Asked
Questions" dates back to 1981-07-13, and there are 11,200 such messages
between then and 1984-01-01. The very first such message connects that
phrase to the acronym "FAQ", contradicting the chronology given for the
birth of that acronym in Wikipedia's entry for "FAQ".
Google Groups' query engine is extremely flaky, but these numbers and
dates are at least suggestive of a different history than the one you
describe.
--
James Kuyper
Joe Pfeiffer
Hmmm.... googling frequently answered questions turns up 5,190,000
hits -- and the first one is a page titled "Frequently Questioned
Answers".
frequently asked questions turns up 222,000,000 hits.
Of course, the claim is that FAQ originally stood for "answered", so
relative use now is pretty much irrelevant.
David Mathog
> 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
J. J. Farrell
> In article <iubgdq$meo$2...@dont-email.me>,
> "J. J. Farrell" <j...@bcs.org.uk> wrote:
>> Michael Press wrote:
>>> Originally FAQ stood for frequently answered questions.
>>> Most people do not seem to know that, even more than
>>> I thought going in. During the transition, some people
>>> attempted to stem the tide.
>> Evidence please. I've known it to be an abbreviation of "Frequently
>> Asked Questions" for several decades, this is the first time I've seen a
>> suggestion that it is "Frequently Answered Questions".
>
> The Wikipedia article on FAQ documents this usage.
Not quite - it claims "Originally the term FAQ referred to the
Frequently Answered Question itself", but with no citation of any sort
to support the claim, and without even identifying the claimant.
Wikipedia does attribute the origin of the acronym FAQ to Eugene Maya on
the SPACE mailing list, and http://www.islandone.org/SpaceDigest/
confirms that Maya originally used the term to mean Frequently Answered
Questions. Since there are records of the term Frequently Asked
Questions which pre-date that by some years, I'd be surprised if that
really was the first time the acronym was used. Quite possible though.
The Wikipedia article doesn't say anything about people attempting to
stem the tide in the supposed change of meaning - where do you get that
from?
> A web search for "frequently answered questions"
> turns up a myriad of sites where the list of frequently
> answered questions is titled "Frequently Answered Questions."
A vastly smaller myriad than that turned up by searching for "frequently
asked questions", neither of which points has any relevance to the
question of which term was the originally intended by the FAQ abbreviation.
Joe Pfeiffer
<snip>
Now, *this* is how one attacks a wikipedia article: not by making a
blanket "wikipedia is unreliable" statment, but by pointing to flaws in
the article's attributions, by pointing out that the article doesn't
quite say what was claimed, and by pointing to better evidence from
elsewhere.
In other words, the way any other argument should be attacked.
DDD
> 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?
>
> 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
In common, you malloc, then you free.
Keith Thompson
Ok, I understand that you're saying that "FAQ" originally stood for
"Frequently Answered Questions". It's not clear whether that's
accurate or not. But whether you're right about that or not,
exactly what point were you trying to make? Does it really matter
whether the 'A' stands for "Asked" or "Answered"?
Eric replied to a question by citing <http://www.c-faq.com/>.
The context implied to me that you were criticizing what Eric did.
Was that your intent? If so, what should he have done? Or was
your point entirely about what the 'A' stands for?
Michael Press
James Kuyper <james...@verizon.net> wrote:
Yes, so they are.
--
Michael Press
Jed
"Billy Mays" <no...@nohow.com> wrote in message
news:iu2sov$3dv$1...@speranza.aioe.org...
>"Should you free your pointers?"
Yes, because, then at least pointers, will be free.