By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
448,671 Members | 1,633 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 448,671 IT Pros & Developers. It's quick & easy.

How to call a constructor explicitly if we want to allocate memory using malloc ?

P: n/a
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?

Oct 18 '06 #1
Share this Question
Share on Google+
13 Replies


P: n/a
shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?
Your create object on the heap with new, not malloc.

A *pA = new A;

You free the memory and object with delete.

delete pA;
Oct 18 '06 #2

P: n/a
shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc".
This will return me the required memory but the object is not
initialized properly as constructor same is not get called ( as per
the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?
Use "placement new" (look it up).

V
--
Please remove capital 'A's when replying by e-mail
I do not respond to top-posted replies, please don't ask
Oct 18 '06 #3

P: n/a

shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?
You can use placement new.

buffer = malloc(/*...*/);
object = new (buffer) MyClass(/*...*/);

But later you have to call the destructor explicitly too:

object->~MyClass();
free(buffer);

Regards,
Bart.

Oct 18 '06 #4

P: n/a
shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?
You can't call a constructor. You can construct an object in
memory you provide (from malloc or anyplace else) with
placement new:

#include <new>

void* memory = malloc(sizeof T);
T* tp = new (memory) T;

Oct 18 '06 #5

P: n/a

Bart wrote:
shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?

You can use placement new.

buffer = malloc(/*...*/);
object = new (buffer) MyClass(/*...*/);

But later you have to call the destructor explicitly too:

object->~MyClass();
free(buffer);

Regards,
Bart.
Is that correct to construct the object on the memory allocated by
malloc? Is that behavior defined in standard C++? I think that the
memory should be allocated by operator new if you want to use placement
new.

Best Regards,
Robin.

Oct 18 '06 #6

P: n/a
zouyongbin wrote:
Bart wrote:
shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).
>
How to call a constructor explicitly if we want to allocate memory
using malloc ?
You can use placement new.

buffer = malloc(/*...*/);
object = new (buffer) MyClass(/*...*/);

But later you have to call the destructor explicitly too:

object->~MyClass();
free(buffer);

Regards,
Bart.

Is that correct to construct the object on the memory allocated by
malloc? Is that behavior defined in standard C++? I think that the
memory should be allocated by operator new if you want to use placement
new.
No. Placement new was invented specifically so you can use your own
memory allocator.

The memory doesn't even have to be "allocated" at all. If you were to
write an OS kernel in C++ you could just construct some object at
whatever physical address you like without ever allocating any memory.

Regards,
Bart.

Oct 18 '06 #7

P: n/a
Ron Natalie wrote:
shsingh wrote:
>I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc". This
will return me the required memory but the object is not initialized
properly as constructor same is not get called ( as per the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?
You can't call a constructor.
Can you call function? or is all that *you* can do writing an expression in
your code and then, if required, the implementation will call a function in
the course of the evaluation of that expression? As far as I know, the
standard does not set linguistic precedence for the using the active voice
(we call a function/constructor) as opposed to the passive voice (a
function/constructor is called) in either case. So do you take a
linguistice license in the case of functions but not in the case of
constructors? If so, why?

You can construct an object in
memory you provide (from malloc or anyplace else) with
placement new:

#include <new>

void* memory = malloc(sizeof T);
T* tp = new (memory) T;

Best

Kai-Uwe Bux

Oct 18 '06 #8

P: n/a
Kai-Uwe Bux wrote:
Ron Natalie wrote:
>shsingh wrote:
>>I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc".
This will return me the required memory but the object is not
initialized properly as constructor same is not get called ( as
per
the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?
You can't call a constructor.

Can you call function? or is all that *you* can do writing an
expression in your code and then, if required, the implementation
will call a function in the course of the evaluation of that
expression? As far as I know, the standard does not set linguistic
precedence for the using the active voice (we call a
function/constructor) as opposed to the passive voice (a
function/constructor is called) in either case. So do you take a
linguistice license in the case of functions but not in the case of
constructors? If so, why?
You can call a function.

You cannot call a constructor directly, because it doesn't have a
name. Most attempts to "call a constructor" will create a temporary of
the class type instead.

class C {};

C::C(); // creates a temporary of class C
In practice, you invoke the constructor as part of executing a new
expression, like placement new.

Bo Persson
Oct 18 '06 #9

P: n/a
Bo Persson wrote:
Kai-Uwe Bux wrote:
>Ron Natalie wrote:
>>shsingh wrote:
I have a class A containing some map as data variables. I creat an
object of class A on heap by allocatiing memory by using "malloc".
This will return me the required memory but the object is not
initialized properly as constructor same is not get called ( as
per
the behavior).

How to call a constructor explicitly if we want to allocate memory
using malloc ?

You can't call a constructor.

Can you call function? or is all that *you* can do writing an
expression in your code and then, if required, the implementation
will call a function in the course of the evaluation of that
expression? As far as I know, the standard does not set linguistic
precedence for the using the active voice (we call a
function/constructor) as opposed to the passive voice (a
function/constructor is called) in either case. So do you take a
linguistice license in the case of functions but not in the case of
constructors? If so, why?

You can call a function.

You cannot call a constructor directly, because it doesn't have a
name. Most attempts to "call a constructor" will create a temporary of
the class type instead.

class C {};

C::C(); // creates a temporary of class C

Come on, can't you spot a troll?

If you create a temporary, doesn't it in fact mean you managed to call
a constructor? Kai-Uwe is just f****ng with you. Now, you've joined
the debacle by introducing your "directly" into the argument. So what
does "you cannot call a constructor directly" actually mean? Does it
mean you *can* call it, or that you *cannot* call it? Drop the
"directly" and explain. Can you? You can't. Directly, I mean. :-)
(now that's what I call a word play)
In practice, you invoke the constructor as part of executing a new
expression, like placement new.
"Invoke", "call"... Isn't it all the same thing? At some point in
the past I grew tired of this debate and now try not to beat this
half-dead horse. Refraining from torturing the poor animal is highly
recommended to all readers.

V
--
Please remove capital 'A's when replying by e-mail
I do not respond to top-posted replies, please don't ask
Oct 18 '06 #10

P: n/a
Bo Persson wrote:
You cannot call a constructor directly, because it doesn't have a
name.
I have read this argument several times. However, I do not understand the
inference: I agree that constructors do not have names. But where in the
standard do you find the other hypothesis needed for the inference, namely
that you need a name to call someting? This additional assumption might be
something that you consider natural because functions are called by using
their names. However, constructor calls do not need to be parallel to
function calls. In my view, it is perfectly natural to regard placement new
as a special syntax provided by the standard for calling constructors.
This syntax does not require constructors to have names, since the class
name is used instead. The compiler then deduces for you which constructors
is to be used (somewhat like it deduces which function you meant to call in
the case of overloaded functions). So again: where in the standard do you
find the provision that something needs to have a name for us to be able to
call it?

Most attempts to "call a constructor" will create a temporary of
the class type instead.

class C {};

C::C(); // creates a temporary of class C
And how does that not call a constructor? If I am not mistaken, an
expression of the form classname( arguments ) is even named an "explicit
constructor call" in the standard.

In practice, you invoke the constructor as part of executing a new
expression, like placement new.
So I can invoke the constructor but I cannot call it? Where in the standard
would I find the distinction between "to invoke" and "to call". As far as I
can see, the two words are both used synonymously in the standard with
regard to constructor calls.
Best

Kai-Uwe Bux
Oct 20 '06 #11

P: n/a
How to call a constructor explicitly if we want to allocate
memory using malloc ?
This thread has been an excellent read for me. "placement new" is a
concept I had not heard of before. Since I do some kernel level
development it is an excellent language feature because you often are
given context buffer space that you overlay a structure onto. Without
placement new, there ar two suboptimal solutions: either avoid
constructors or put a pointer to the object (instead of the object
itself) in the context buffer and new/delete it.

If there are any other concepts that might especially be useful in
kernel level development, please mention.

Oct 21 '06 #12

P: n/a
* Kai-Uwe Bux:
[about "can't call constructor" urban myth]

I recently ran into yet another formulation in the standard that uses
the "call" terminology, namely 1.9/15, about an initializer like

Foo object( arg1, arg2, arg3 ... );

"the resulting construct is a function call upon a constructor function,
with expression-list as an argument list; such a function call is a
full-expression".

Can't be very much more clear than that, I think. ;-)

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
Oct 21 '06 #13

P: n/a
44****@email.com wrote:
How to call a constructor explicitly if we want to allocate
memory using malloc ?

This thread has been an excellent read for me. "placement new" is a
concept I had not heard of before. Since I do some kernel level
development it is an excellent language feature because you often are
given context buffer space that you overlay a structure onto. Without
placement new, there ar two suboptimal solutions: either avoid
constructors or put a pointer to the object (instead of the object
itself) in the context buffer and new/delete it.
There are many non-portable ways that you could manage objects at a
fixed memory location in a kernel (finding the address of the
constructor, relocating the stack, telling the linker where to
construct the objects) but there are at least a few ways that you could
do it using standard C++ features:

- placement new
- overloaded new operator
- pointers to fixed addresses (works for POD only)
If there are any other concepts that might especially be useful in
kernel level development, please mention.
As mentionned above, you can overload new and delete. You can also
provide your own allocator to the standard containers. Such things
might be useful in a kernel implementation. However, some features such
as exceptions and RTTI may be unavailable or may need special
implementation.

The following link may be interesting:
http://netlab.ru.is/exception/LinuxCXX.shtml

Regards,
Bart.

Oct 21 '06 #14

This discussion thread is closed

Replies have been disabled for this discussion.