how free() works

[riva]

We merely pass the pointer to dynamically allocated variable to free()
function.
How does it decide about the number of bytes to be freed?

[Zara]

On Tue, 25 Dec 2007 22:16:22 -0800 (PST), riva <ra*********@gmail.com>
wrote:

>We merely pass the pointer to dynamically allocated variable to free()
function.
How does it decide about the number of bytes to be freed?

It is implementation dependent. Usually, the chunk of bytes realy
assigned with malloc is at least sizeof(size_t) bytes larger then the
request. These extra bytes are used to store the size of the chunk,
and they are 'before' the pointer returned. But please remember thet
each compiler is free to do as it likes, so thi is no rule of thumb.

Best regards,

Zara

[Vinay]

Guys..i want to do the following!

I want to edit a text file through my C program.

The text file is as follows
1
2
3
4

i want to change the contents to

1
2
100
3
4

Thanks in advance..

[Chris Dollin]

riva wrote:

We merely pass the pointer to dynamically allocated variable to free()
function.
How does it decide about the number of bytes to be freed?

`malloc` and `free` are friends and have no secrets from one
another.

Different implementations, like different friends, have different
secrets.

--
Conspiracy Hedgehog
Meaning precedes definition.

[Keith Thompson]

riva <ra*********@gmail.comwrites:

We merely pass the pointer to dynamically allocated variable to free()
function.
How does it decide about the number of bytes to be freed?

The comp.lang.c FAQ is at <http://www.c-faq.com/>. You've just asked
question 7.26.

--
Keith Thompson (The_Other_Keith) <ks***@mib.org>
Looking for software development work in the San Diego area.
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

[Malcolm McLean]

"Vinay" <vi*****@gmail.comwrote in message

Guys..i want to do the following!

I want to edit a text file through my C program.

The text file is as follows
1
2
3
4

i want to change the contents to

1
2
100
3
4

Thanks in advance..

Please start a new thread for a new topic.

The problem hasn't been very well specified.
You can achieve the output demanded simply by calling fopen() fprintf() and
fclose(), with the name of the file you wish to overwrite.
However this is unlikely to be what you really want. You'll have to ask for
a slightly better formal spec. Do you want to insert an arbitrary
user-specified number in the second line, of arbitrary input text, for
example?

--
Free games and programming goodies.
http://www.personal.leeds.ac.uk/~bgy1mm

[Richard]

Chris Dollin <[e-mail]:

riva wrote:

>We merely pass the pointer to dynamically allocated variable to free()
function.
How does it decide about the number of bytes to be freed?

`malloc` and `free` are friends and have no secrets from one
another.

Different implementations, like different friends, have different
secrets.

Nice explanation and analogy.

[CBFalconer]

riva wrote:

>
We merely pass the pointer to dynamically allocated variable to
free() function. How does it decide about the number of bytes to
be freed?

Raw black magic.

--
Merry Christmas, Happy Hanukah, Happy New Year
Joyeux Noel, Bonne Annee, Frohe Weihnachten
Chuck F (cbfalconer at maineline dot net)
<http://cbfalconer.home.att.net>
--
Posted via a free Usenet account from http://www.teranews.com

[Flash Gordon]

CBFalconer wrote, On 26/12/07 21:11:

riva wrote:

>We merely pass the pointer to dynamically allocated variable to
free() function. How does it decide about the number of bytes to
be freed?

Raw black magic.

Nope. On mine the black magic has been grilled under a low heat.
--
Flash Gordon

[CBFalconer]

Flash Gordon wrote:

CBFalconer wrote, On 26/12/07 21:11:

>riva wrote:

>>We merely pass the pointer to dynamically allocated variable to
free() function. How does it decide about the number of bytes to
be freed?

Raw black magic.

Nope. On mine the black magic has been grilled under a low heat.

Don't you find that that allows the basic greasy black material to
leak out, so that heavily realloced buffers tend to fail to free
abandoned buffers, resulting in memory leaks?

Code such as the following can be used to detect this sort of
'black leak':

size_t ct
char *p, *t;

for (p = NULL, ct = 0; ++ct;) {
if (t = realloc(p, ct)) p = t;
else break;
}
if (t == p) /* no realloc failure occured */
analyzemem(p); /* system dependant code */
}
else failanalyze(p); /* also system dependant */

--
Merry Christmas, Happy Hanukah, Happy New Year
Joyeux Noel, Bonne Annee, Frohe Weihnachten
Chuck F (cbfalconer at maineline dot net)
<http://cbfalconer.home.att.net>

--
Posted via a free Usenet account from http://www.teranews.com

[Flash Gordon]

CBFalconer wrote, On 27/12/07 21:14:

Flash Gordon wrote:

>CBFalconer wrote, On 26/12/07 21:11:

>>riva wrote:

We merely pass the pointer to dynamically allocated variable to
free() function. How does it decide about the number of bytes to
be freed?
Raw black magic.

Nope. On mine the black magic has been grilled under a low heat.

Don't you find that that allows the basic greasy black material to
leak out, so that heavily realloced buffers tend to fail to free
abandoned buffers, resulting in memory leaks?

No, since the black magic is grilled *before* it is used for
malloc/realloc/free.

Code such as the following can be used to detect this sort of
'black leak':

size_t ct
char *p, *t;

for (p = NULL, ct = 0; ++ct;) {
if (t = realloc(p, ct)) p = t;
else break;
}

My optimiser replaces this entire loop with:
p = __malloc_largest_possible_block(&ct);
t = 0;
so no reallocing actually occurs.

if (t == p) /* no realloc failure occured */
analyzemem(p); /* system dependant code */
}
else failanalyze(p); /* also system dependant */

--
Flash Gordon