Stacks and heaps

[Sathyaish]

I've searched Google and found a few, but I am not so satisfied. Any
good reading on "stacks and heaps" about how and when memory is
allocated from the heap?

 

[Jack Klein]

I've searched Google and found a few, but I am not so satisfied. Any
good reading on "stacks and heaps" about how and when memory is
allocated from the heap?

The C standard does not specify mechanisms like stacks and heaps.
Memory has three types of storage duration in C programs, static,
automatic, and dynamic.

When, how, and why an implementation provides and manages these types
of memory is up to the implementation. So if you are interested in
how a specific compiler manages memory, you need to ask in a compiler
specific support group.

 

[Dan Pop]

The C standard does not specify mechanisms like stacks and heaps.

These are data structures, not mechanisms.

Memory has three types of storage duration in C programs, static,
automatic, and dynamic.

When, how, and why an implementation provides and manages these types
of memory is up to the implementation. So if you are interested in
how a specific compiler manages memory, you need to ask in a compiler
specific support group.

There is nothing compiler specific in understanding how these data
structures can be used to implement automatic and dynamic memory
allocation.

Dan

 

[Thomas Matthews]

I've searched Google and found a few, but I am not so satisfied. Any
good reading on "stacks and heaps" about how and when memory is
allocated from the heap?

Often times, compilers use a heap for managing dynamic memory.
Dynamic memory is memory that is set aside for the program to
use during run-time.

Most of the time, programs use the *alloc() family of functions
to allocate from dynamic memory. The memory stays allocated until
either the program deallocates the memory via free or the
program is terminated (and the memory is re-used by the operating
system).

Dynamic memory often is used when the amount cannot be determined
during compile time, or when the program needs more memory than
the compiler can allocate during compile time (or program load
time).

Linked lists are a classic example of using dynamic memory
allocation.

--
Thomas Matthews

C++ newsgroup welcome message:
http://www.slack.net/~shiva/welcome.txt
C++ Faq: http://www.parashift.com/c++-faq-lite
C Faq: http://www.eskimo.com/~scs/c-faq/top.html
alt.comp.lang.learn.c-c++ faq:
http://www.raos.demon.uk/acllc-c++/faq.html
Other sites:
http://www.josuttis.com -- C++ STL Library book

 

[Sathyaish]

Thanks, Thomas and others. I want some more rigorous reading on this.
For instance, I could easily understand what Thomas Matthews said
because I knew it earlier plus I have some background of C and C++.
Where can I get more copious information on the internal working of
the heap?

PS: Thomas, haven't I seen you on JoS?

 

[Thomas Matthews]

Thanks, Thomas and others. I want some more rigorous reading on this.
For instance, I could easily understand what Thomas Matthews said
because I knew it earlier plus I have some background of C and C++.
Where can I get more copious information on the internal working of
the heap?

The heap is a tool (data structure) used by some compiler vendors to
manage dynamic memory. An implementation is not required to have
a heap in order to be compliant with the standard.

The question now is whether you want copious information about:
1. Heap data structures
2. Memory allocation methods (schemes)
3. Compiler Theory.
I suggest using your favorite web search engine with the appropriate
keywords and see what you can find. I took and older route and took
classes in the above at the University. Many Universities give the
Computer Science people a Compiler to write as their final project.

As far as the C language goes, you can do three things with
dynamic memory:
1. Allocate from it.
2. Reallocate from it.
2. Deallocate objects to it (i.e. return or free already
allocated items).
Kind of simple and boring. You allocate memory as you need it
and dispose of the objects when you are finished with them.
Anything beyond this, such as alignment and memory collection,
is beyond the scope of this newsgroup.

PS: Thomas, haven't I seen you on JoS?

What is JoS?
I remember playing with an operating system called JOSS, but
that was a long time ago.

If you know me, you might remember that I really hate acroynyms.
Just as a picture is worth a thousand words, one never really
knows which of those thousand words the author/artist was meaning
to convey. I prefer to be explicit. So, WTF is JoS?


--
Thomas Matthews

C++ newsgroup welcome message:
http://www.slack.net/~shiva/welcome.txt
C++ Faq: http://www.parashift.com/c++-faq-lite
C Faq: http://www.eskimo.com/~scs/c-faq/top.html
alt.comp.lang.learn.c-c++ faq:
http://www.raos.demon.uk/acllc-c++/faq.html
Other sites:
http://www.josuttis.com -- C++ STL Library book

 

[Sathyaish]

That answers it. The Joel On Software site, a wonderful site about
software development issues, is more fondly known by its members as
JoS. Thanks for the info. I wish I'd gone to a university for a
Computer Science degree when my parents couldn't afford to send me.

Regards,
Sathyaish Chakravarthy.