473,386 Members | 1,973 Online
Bytes | Software Development & Data Engineering Community
Post Job

Home Posts Topics Members FAQ

Join Bytes to post your question to a community of 473,386 software developers and data experts.

How does "delete [] <pointer>" work?

Hi everyone,

Im just starting out with C++, and am curious to know how "delete []
<pointer>", knows about the number of memory locations to free.

I read somewhere that delete frees up space assigned to <pointerby
"new". Does "new" create a list of pointer names and the size of the
memory array they point to? I also read that some compilers may store
the number of consec mem locations a pointer points to, just before the
first data element. Is that correct?

Also, if i used malloc() to make <pointerpoint to a number of memory
locations, will calling "delete [] <pointer>" still work? if yes, how
does delete here know the number of memory locations to free? does the
malloc() implementation in C++ also create a pointer/memsize list like
"new"?

And lastly, if i made another pointer <pointer2point to the same
location as <pointer>, would calling "delete [] <pointer2>" still free
up the correct number of locations?

Thanks in advance for answering my deluge of questions!
regards,
Mayank

Aug 20 '06 #1
5 8050
mkaushik wrote:
Hi everyone,

Im just starting out with C++, and am curious to know how "delete []
<pointer>", knows about the number of memory locations to free.
The answer is implementation dependent.
I read somewhere that delete frees up space assigned to <pointerby
"new". Does "new" create a list of pointer names and the size of the
memory array they point to? I also read that some compilers may store
the number of consec mem locations a pointer points to, just before the
first data element. Is that correct?
In general, that's how it works, but you can't rely on it. Each
implementation has its own methods.
>
Also, if i used malloc() to make <pointerpoint to a number of memory
locations, will calling "delete [] <pointer>" still work?
No. If you malloc() it, free() it. If you "new" it, "delete" it. If
you "new[]" it, "delete[]" it. It's that simple.

if yes, how
does delete here know the number of memory locations to free? does the
malloc() implementation in C++ also create a pointer/memsize list like
"new"?
Many implementations of operator new call malloc behind the scenes, but
it is not required. Again, it's implementation dependent. Besides, you
can't delete[] memory allocated with malloc.
And lastly, if i made another pointer <pointer2point to the same
location as <pointer>, would calling "delete [] <pointer2>" still free
up the correct number of locations?
int* p = new int[30];
int* p2 = p;
delete[] p2;

Is valid, but dangerous, as p now points at unallocated memory.
Accessing anything through p (or p2) is undefined behavior.

Aug 20 '06 #2
In article <11**********************@m79g2000cwm.googlegroups .com>,
pr***************@yahoo.com says...
Hi everyone,

Im just starting out with C++, and am curious to know how "delete []
<pointer>", knows about the number of memory locations to free.
It's required to do so, but exactly how it does so is up to the
implementation.
I read somewhere that delete frees up space assigned to <pointerby
"new". Does "new" create a list of pointer names and the size of the
memory array they point to? I also read that some compilers may store
the number of consec mem locations a pointer points to, just before the
first data element. Is that correct?
In both cases, the answer is a definite maybe -- it can keep track of
all of the data in one place, or it can put data in each node as it's
allocated, or whatver else it prefers.
Also, if i used malloc() to make <pointerpoint to a number of memory
locations, will calling "delete [] <pointer>" still work?
If you use malloc to allocate some memory and then use delete on that
memory, the results are undefined -- it might work for some sufficiently
loose definition of "work". Then again, it might simply crash and burn,
or it might do just about anything else.
if yes, how
does delete here know the number of memory locations to free? does the
malloc() implementation in C++ also create a pointer/memsize list like
"new"?
It could, but it's not required to do so -- allocating memory with
malloc and then attempting to delete it isn't required to work at all.
And lastly, if i made another pointer <pointer2point to the same
location as <pointer>, would calling "delete [] <pointer2>" still free
up the correct number of locations?
If I understand your question correctly, yes. What you get back from new
is an address. You have to pass that address to delete exactly once --
but the pointer you store the address is makes no difference at all.

To summarize: if you use new to allocate some memory, you must use
delete exactly once on the address you get from new. If you use new[] to
allocate the memory, you must use delete[] exactly once on that address.
If you allocate some memory with malloc, you must call free exactly once
with that address malloc returned.

Beyond that, exactly how each of these accomplishes what it's supposed
to is up to the implementer. When you do something like:

struct whatever {
whatever() { std::cout << "ctor\n"; }
~whatever() { std::cout << "dtor\n"; }
};

whatever *w = new whatever[10];
// ... other stuff here

delete [] whatever;

The 'new' is reuqired to use whatever's ctor to create each of the ten
objects, and the delete is required to use whatever's dtor to destroy
each of those ten objects. How it knows there are ten objects to destroy
is entirely up to it. It might allocate some extra memory and store it
along with the object. It might have a separate table of sizes of
objects. It's even possible that it could (for one example) use the
exact value of the pointer to tell it what it needs -- assuming, for the
moment, a CPU that had no alignment requirements, it could (for one
example) make some bits in the address it assigns equal to the size of
the array that was allocated. Obviously this has the potential for
wasting some address space, but on (for one example) a machine with 64-
bit addressing, that's not likely to be a problem at least in the near
future.

--
Later,
Jerry.

The universe is a figment of its own imagination.
Aug 20 '06 #3
In article <MP************************@news.sunsite.dk>,
jc*****@taeus.com says...

[ ... ]
whatever *w = new whatever[10];
// ... other stuff here

delete [] whatever;
Oops -- that should (of course) be:

delete [] w;

Sorry 'bout that.

--
Later,
Jerry.

The universe is a figment of its own imagination.
Aug 20 '06 #4
In article <11**********************@m79g2000cwm.googlegroups .com>,
mkaushik <pr***************@yahoo.comwrote:
>Im just starting out with C++, and am curious to know how "delete []
<pointer>", knows about the number of memory locations to free.

I read somewhere that delete frees up space assigned to <pointerby
"new". Does "new" create a list of pointer names and the size of the
memory array they point to? I also read that some compilers may store
the number of consec mem locations a pointer points to, just before the
first data element. Is that correct?
Exactly how it works is implementation defined, so yes, that's
one allowed possibility as I recall.
>Also, if i used malloc() to make <pointerpoint to a number of memory
locations, will calling "delete [] <pointer>" still work? if yes, how
does delete here know the number of memory locations to free? does the
malloc() implementation in C++ also create a pointer/memsize list like
"new"?
It may work, but generally a bad idea; most likely it'll crash,
or appear to work only for something seemingly unrelated to fail later.
Normally you want to stay in the same deallocation family that you
allocated from. So malloc/free, new/delete, new[]/delete[],
class new/delete / new[]/delete[] respectively. "placement new"
has its own concerns too.
>And lastly, if i made another pointer <pointer2point to the same
location as <pointer>, would calling "delete [] <pointer2>" still free
up the correct number of locations?
Absolutely. Just make sure your code doesn't use either after
that point though because they'd be invalid pointers unless
reassigned, etc.
--
Greg Comeau / 20 years of Comeauity! Intel Mac Port now in alpha!
Comeau C/C++ ONLINE == http://www.comeaucomputing.com/tryitout
World Class Compilers: Breathtaking C++, Amazing C99, Fabulous C90.
Comeau C/C++ with Dinkumware's Libraries... Have you tried it?
Aug 20 '06 #5
mkaushik wrote:
I read somewhere that delete frees up space assigned to <pointerby
"new". Does "new" create a list of pointer names and the size of the
memory array they point to? I also read that some compilers may store
the number of consec mem locations a pointer points to, just before the
first data element. Is that correct?
See also
http://www.parashift.com/c++-faq-lit...html#faq-16.14

Phil
Aug 21 '06 #6

This thread has been closed and replies have been disabled. Please start a new discussion.

Similar topics

22
by: Dr Duck | last post by:
GDay all, Something seems odd to me.... I wrote a simple C# function public void bind(ref object a, ref object b, bool atob) { if(atob) b = a; else
3
by: Piotre Ugrumov | last post by:
I have done the overload on the operator >> and << in the class Attore. These 2 overload work correctly. I have done the overload of the same overload in the class Film. The class film ha inside...
10
by: Eric-Sebastien Lachance | last post by:
Hey there, I decided to just create a 100% height and width div that filled the space over a background header image, and add an onclick event to redirect to the my index... Doesn't seem to work...
20
by: Oliver S. | last post by:
I was often annoyed about the performance of std::basic-string because of complex memory-allocators working in the background to found the work of std::allocator<T>. Basically, the performance...
3
by: Evgeny | last post by:
Hi, all! I didn't find yet solution for this problem! Somebody knows where is a catch? Looks like "operator =" or copy constructor not implemented in one of internal templates.... Thanks in...
16
by: call_me_anything | last post by:
why is the following not allowed : vector <Base *vec_of_base; vector <Derived *vec_of_derived; vec_of_base = vec_of_derived; Note : The following is allowed :
232
by: robert maas, see http://tinyurl.com/uh3t | last post by:
I'm working on examples of programming in several languages, all (except PHP) running under CGI so that I can show both the source files and the actually running of the examples online. The first...
1
tekninja
by: tekninja | last post by:
I wanted to write a function to take the tedium out of dynamically determining the total number of elements in an array. Within the main method this is easy as I simply use: sizeof( <array name>...
43
by: john | last post by:
Hi, in TC++PL 3 on pages 674-675 it is mentioned: "Maybe your first idea for a two-dimensional vector was something like this: class Matrix { valarray< valarray<doublev; public: // ... };
0
by: Charles Arthur | last post by:
How do i turn on java script on a villaon, callus and itel keypad mobile phone
0
by: ryjfgjl | last post by:
If we have dozens or hundreds of excel to import into the database, if we use the excel import function provided by database editors such as navicat, it will be extremely tedious and time-consuming...
0
by: ryjfgjl | last post by:
In our work, we often receive Excel tables with data in the same format. If we want to analyze these data, it can be difficult to analyze them because the data is spread across multiple Excel files...
0
by: emmanuelkatto | last post by:
Hi All, I am Emmanuel katto from Uganda. I want to ask what challenges you've faced while migrating a website to cloud. Please let me know. Thanks! Emmanuel
0
BarryA
by: BarryA | last post by:
What are the essential steps and strategies outlined in the Data Structures and Algorithms (DSA) roadmap for aspiring data scientists? How can individuals effectively utilize this roadmap to progress...
0
by: Hystou | last post by:
There are some requirements for setting up RAID: 1. The motherboard and BIOS support RAID configuration. 2. The motherboard has 2 or more available SATA protocol SSD/HDD slots (including MSATA, M.2...
0
marktang
by: marktang | last post by:
ONU (Optical Network Unit) is one of the key components for providing high-speed Internet services. Its primary function is to act as an endpoint device located at the user's premises. However,...
0
by: Hystou | last post by:
Most computers default to English, but sometimes we require a different language, especially when relocating. Forgot to request a specific language before your computer shipped? No problem! You can...
0
Oralloy
by: Oralloy | last post by:
Hello folks, I am unable to find appropriate documentation on the type promotion of bit-fields when using the generalised comparison operator "<=>". The problem is that using the GNU compilers,...

By using Bytes.com and it's services, you agree to our Privacy Policy and Terms of Use.

To disable or enable advertisements and analytics tracking please visit the manage ads & tracking page.