Can realloc(p,0) return NULL when p is non-NULL and memory aplenty?

F

Francois Grieu

When running the following code under MinGW, I get
realloc(p,0) returned NULL
Is that a non-conformance?

TIA,
Francois Grieu

#include <stdio.h>
#include <stdlib.h>

int main(void)
{
void *p;
p = malloc(0);
if (p==NULL)
puts("malloc(0) returned NULL");
else
{
p = realloc(p,0);
if (p==NULL)
puts("realloc(p,0) returned NULL");
else
puts("realloc(p,0) did not return NULL");
}
return 0;
}
 
J

Joachim Schmitz

Francois said:
When running the following code under MinGW, I get
realloc(p,0) returned NULL
Is that a non-conformance?
Whether malloc(0) returns NULL or a pointerr to 0 bytes is implementation
defined
I think the same is true for realloc(p, 0)

Bye, Jojo
 
W

Walter Roberson

When running the following code under MinGW, I get
realloc(p,0) returned NULL
Is that a non-conformance?

No, it is conformance, and returning non-NULL would be non-conformance.

C89 4.10.3.4 The realloc Funciton

If size is zero and ptr is not a null pointer, the object
it points to is freed.

Returns

The realloc function returns either a null pointer or a
pointer to the possibly moved allocated space.


In the case of zero as the size, there is no more allocated space
(because zero size requires freeing), so the other branch of the
'either' kicks in, requiring realloc to return a null pointer
for this case.
 
J

Joachim Schmitz

Walter said:
No, it is conformance, and returning non-NULL would be
non-conformance.

C89 4.10.3.4 The realloc Funciton

If size is zero and ptr is not a null pointer, the object
it points to is freed.

Returns

The realloc function returns either a null pointer or a
pointer to the possibly moved allocated space.


In the case of zero as the size, there is no more allocated space
(because zero size requires freeing), so the other branch of the
'either' kicks in, requiring realloc to return a null pointer
for this case.
From n1256:
7.20.3 Memory management functions

1 The order and contiguity of storage allocated by successive calls to the
calloc,
malloc, and realloc functions is unspecified. The pointer returned if the
allocation
succeeds is suitably aligned so that it may be assigned to a pointer to any
type of object
and then used to access such an object or an array of such objects in the
space allocated
(until the space is explicitly deallocated). The lifetime of an allocated
object extends
from the allocation until the deallocation. Each such allocation shall yield
a pointer to an
object disjoint from any other object. The pointer returned points to the
start (lowest byte
address) of the allocated space. If the space cannot be allocated, a null
pointer is
returned. If the size of the space requested is zero, the behavior is
implementationdefined:
either a null pointer is returned, or the behavior is as if the size were
some
nonzero value, except that the returned pointer shall not be used to access
an object.

....

7.20.3.4 The realloc function
....
3 If ptr is a null pointer, the realloc function behaves like the malloc
function for the
specified size



So it surely seems implementation defined to me.



Bye, Jojo
 
J

Joachim Schmitz

Joachim said:
From n1256:
7.20.3 Memory management functions

1 The order and contiguity of storage allocated by successive calls
to the calloc,
malloc, and realloc functions is unspecified. The pointer returned if
the allocation
succeeds is suitably aligned so that it may be assigned to a pointer
to any type of object
and then used to access such an object or an array of such objects in
the space allocated
(until the space is explicitly deallocated). The lifetime of an
allocated object extends
from the allocation until the deallocation. Each such allocation
shall yield a pointer to an
object disjoint from any other object. The pointer returned points to
the start (lowest byte
address) of the allocated space. If the space cannot be allocated, a
null pointer is
returned. If the size of the space requested is zero, the behavior is
implementationdefined:
either a null pointer is returned, or the behavior is as if the size
were some
nonzero value, except that the returned pointer shall not be used to
access an object.

...

7.20.3.4 The realloc function
...
3 If ptr is a null pointer, the realloc function behaves like the
malloc function for the
specified size
Oops, that section is irrelevant...
So it surely seems implementation defined to me.
But from the 1st section I think this still holds true

Bye, Jojo
 
J

Joachim Schmitz

Joachim said:
Oops, that section is irrelevant...

But from the 1st section I think this still holds true
Esp. as the section Walter quoted isn't in n1256 (anymore?)

Bye, Jojo
 
H

Harald van Dijk

When running the following code under MinGW, I get realloc(p,0) returned
NULL
Is that a non-conformance?

No, it is conformance, and returning non-NULL would be non-conformance.

[C89 citation snipped]

While you're not at all wrong, please keep in mind that this is one of the
areas in which C99 differs from the previous standard. In C99, it's
unspecified whether realloc(p, 0) returns a null pointer, but if it
returns a null pointer, then p is *not* freed.
 
J

Joachim Schmitz

Harald said:
When running the following code under MinGW, I get realloc(p,0)
returned NULL
Is that a non-conformance?

No, it is conformance, and returning non-NULL would be
non-conformance.

[C89 citation snipped]

While you're not at all wrong, please keep in mind that this is one
of the areas in which C99 differs from the previous standard. In C99,
it's unspecified whether realloc(p, 0) returns a null pointer, but if
it returns a null pointer, then p is *not* freed.
I don't read it mlike this

The realloc function returns a pointer to the new object (which may have the
same
value as a pointer to the old object), or a null pointer if the new object
could not be
allocated.

allocating 0 bytes can't be too difficult ;-), so no compelling reason to
return NULL

Bye, Jojo
 
W

Walter Roberson

When running the following code under MinGW, I get realloc(p,0) returned
NULL
Is that a non-conformance?
No, it is conformance, and returning non-NULL would be non-conformance.
[C89 citation snipped]
While you're not at all wrong, please keep in mind that this is one of the
areas in which C99 differs from the previous standard. In C99, it's
unspecified whether realloc(p, 0) returns a null pointer, but if it
returns a null pointer, then p is *not* freed.

Good point -- but as best I -recall- MinGW is not C99.
 
F

Francois Grieu

No, it is conformance, and returning non-NULL would be non-conformance.

I disagree (at least) about the second member of that answer under
C99. IMHO realloc(p,0) is allowed to return the same thing as
malloc(0), which is not required to return NULL.

I reason that when p is a valid non-NULL pointer to a pointer returned
by malloc() and memory is aplenty, realloc(p,0) must return a pointer
to a new object that has the size 0, which is the same thing a what
malloc(0) should return, which is not required to be NULL, and is not
NULL in many implementations, including that one I'm strugling with.
C89 4.10.3.4 The realloc Function

If size is zero and ptr is not a null pointer, the object
it points to is freed.

Returns

The realloc function returns either a null pointer or a
pointer to the possibly moved allocated space.

In the case of zero as the size, there is no more allocated space
(because zero size requires freeing), so the other branch of the
'either' kicks in, requiring realloc to return a null pointer
for this case.


C99 says {
void *realloc(void *ptr, size_t size);

The realloc function deallocates the old object pointed to by ptr and
returns a pointer to a new object that has the size specified by size.
The contents of the new object shall be the same as that of the old
object prior to deallocation, up to the lesser of the new and old
sizes. Any bytes in the new object beyond the size of the old object
have indeterminate values.

If ptr is a null pointer, the realloc function behaves like the malloc
function for the specified size. Otherwise, if ptr does not match a
pointer earlier returned by the calloc, malloc, or realloc function,
or if the space has been deallocated by a call to the free or realloc
function, the behavior is undefined. If memory for the new object
cannot be allocated, the old object is not deallocated and its value
is unchanged.

The realloc function returns a pointer to the new object (which may
have the same value as a pointer to the old object), or a null pointer
if the new object could not be allocated.
}




I get the same result "realloc(p,0) returned NULL" with the source
changed to the deeper test:

#include <stdio.h>
#include <stdlib.h>

int main(void)
{
void *p = malloc(1);
if (p==NULL)
puts("Memory is very low"); // does not occur
if (malloc(0)==NULL)
puts("malloc(0) returned NULL"); // does not occur
if (realloc(NULL,0)==NULL)
puts("realloc(NULL,0) returned NULL"); // does not occur
if (realloc(p,0)==NULL)
puts("realloc(p,0) returned NULL"); // occurs ?!
if (malloc(10000)==NULL)
puts("Memory was rather low"); // does not occur
}


So I get an implementation of realloc(ptr,0) that turns a NULL ptr to
a non-NULL result, and non-NULL ptr to NULL result; which is at the
very least a surprise, non-orthogonal, and error-prone. In my case a
perfectly sound-looking "read the whole file in memory, in several
chuncks as necessary" function failed for an empty file, claiming
there is no memory.

Francois Grieu
 
H

Harald van Dijk

Harald said:
When running the following code under MinGW, I get realloc(p,0)
returned NULL
Is that a non-conformance?

No, it is conformance, and returning non-NULL would be
non-conformance.

[C89 citation snipped]

While you're not at all wrong, please keep in mind that this is one of
the areas in which C99 differs from the previous standard. In C99, it's
unspecified whether realloc(p, 0) returns a null pointer, but if it
returns a null pointer, then p is *not* freed.
I don't read it mlike this

The realloc function returns a pointer to the new object (which may have
the same
value as a pointer to the old object), or a null pointer if the new
object could not be
allocated.

7.20.3p1 applies to all allocation functions.
"If the size of the space requested is zero, the behavior is
implementation-defined: either a null pointer is returned, or the
behavior is as if the size were some nonzero value, except that the
returned pointer shall not be used to access an object."

This allows realloc(p, 0) to unconditionally fail (since the text you
quoted specifies that a null pointer return value signifies failure for
realloc).
allocating 0 bytes can't be too difficult ;-)

<nit>realloc(p, 0), if it succeeded, hasn't allocated 0 bytes. It has
allocated one or more bytes, plus whatever information free needs to give
it back later.</nit>

Anyway, even if it weren't specifically allowed to always fail, it could
still for example fail on implementations where if the size is small,
*alloc returns pointers to pre-allocated buckets depending on the
requested size.
 
C

CBFalconer

Walter said:
No, it is conformance, and returning non-NULL would be non-conformance.
.... snip C89 quote ...

C99 quote:

7.20.3 Memory management functions

[#1] The order and contiguity of storage allocated by
successive calls to the calloc, malloc, and realloc
functions is unspecified. The pointer returned if the
allocation succeeds is suitably aligned so that it may be
assigned to a pointer to any type of object and then used to
access such an object or an array of such objects in the
space allocated (until the space is explicitly freed or
reallocated). Each such allocation shall yield a pointer to
an object disjoint from any other object. The pointer
returned points to the start (lowest byte address) of the
allocated space. If the space cannot be allocated, a null
pointer is returned. If the size of the space requested is <
zero, the behavior is implementation-defined: either a null <
pointer is returned, or the behavior is as if the size were <
some nonzero value, except that the returned pointer shall <
not be used to access an object. The value of a pointer <
that refers to freed space is indeterminate. <

Note that either a NULL or an individual pointer can be returned.
The pointer to a zero size space can't be dereferenced. You also
can't positively tell a malloc system error from success. So I
advise always allocating (or reallocing) 1 or more bytes.
 
F

Falcon Kirtaran

Francois said:
When running the following code under MinGW, I get
realloc(p,0) returned NULL
Is that a non-conformance?

TIA,
Francois Grieu

#include <stdio.h>
#include <stdlib.h>

int main(void)
{
void *p;
p = malloc(0);
if (p==NULL)
puts("malloc(0) returned NULL");
else
{
p = realloc(p,0);
if (p==NULL)
puts("realloc(p,0) returned NULL");
else
puts("realloc(p,0) did not return NULL");
}
return 0;
}

In the UNIX manpage for the function, you will find the following text:

realloc() returns a pointer to the newly allocated memory, which is
suitably aligned for any kind of variable and may be different from
ptr, or NULL if the request fails. If size was equal to 0, either
NULL or a pointer suitable to be passed to free() is returned. If
realloc() fails the original block is left untouched; it is not freed or
moved.

Now, it really does beg the question why you are trying to realloc() the
void * to be 0 bytes long. The function you should use for this is
free() (and you will find in the same manpage that realloc(x, 0) and
free(x) are equivalent.

Nevertheless, that behaviour is defined and correct.
 
J

Joachim Schmitz

CBFalconer wrote:
You also can't positively tell a malloc system error from success.
Probably not with standrd C, but AFAIK POSIX reqires malloc to set errno on
failure.
So if malloc() return NULL, lookup errno, if that is 0, no failure occured.

Bye, Jojo
 
M

Micah Cowan

CBFalconer said:
Note that either a NULL or an individual pointer can be returned.
The pointer to a zero size space can't be dereferenced. You also
can't positively tell a malloc system error from success. So I
advise always allocating (or reallocing) 1 or more bytes.

While that's true, I'm not sure I see what practical use it would make
to be able to determine the difference between NULL returned because
of a system error, and NULL returned because you asked for either that
or an unusable pointer.

That being said, I agree about reallocing 1 or more bytes, since doing
0 without knowing that your pointer will be freed is pretty useless.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

No members online now.

Forum statistics

Threads
473,769
Messages
2,569,580
Members
45,054
Latest member
TrimKetoBoost

Latest Threads

Top