Stack overflows

[Daz]

Hi everyone!

It is indeed, once again time for me to ask another stupid question.

I have been searching around on the net for quite a while, and can't
find anything that explains 'exactly' what a stack overflow is, or how
it's detected. I know that it's leterally putting too many objects on
the stack, but how many objects can you have on the stack, or do OSs
only allow a certain number of recursive calls to functions?

I apologise in advance if I am missing something here.

Regards

Daz

[Phlip]

Re: Stack overflows

Daz wrote:
[color=blue]
> It is indeed, once again time for me to ask another stupid question.
>
> I have been searching around on the net for quite a while, and can't
> find anything that explains 'exactly' what a stack overflow is, or how
> it's detected. I know that it's leterally putting too many objects on
> the stack, but how many objects can you have on the stack, or do OSs
> only allow a certain number of recursive calls to functions?
>
> I apologise in advance if I am missing something here.[/color]

You have been posting here long enough to notice all the screams over
off-topic questions, right?

The limit of your stack is implementation-defined, and the behavior when you
overflow it is either implementation-defined or undefined.

You could write a program that calls void foo() { foo(); } and run it to see
what happens. It might be a trappable error, or might be a GPF. All
completely off-topic, because only an on-topic newsgroup can tell you how
reliable the error will be, and how preventable, how recoverable, etc.

--
Phlip
http://c2.com/cgi/wiki?ZeekLand <-- NOT a blog!!!

[Daz]

Re: Stack overflows


Phlip wrote:[color=blue]
> Daz wrote:
> You have been posting here long enough to notice all the screams over
> off-topic questions, right?
>
> The limit of your stack is implementation-defined, and the behavior when you
> overflow it is either implementation-defined or undefined.
>
> You could write a program that calls void foo() { foo(); } and run it to see
> what happens. It might be a trappable error, or might be a GPF. All
> completely off-topic, because only an on-topic newsgroup can tell you how
> reliable the error will be, and how preventable, how recoverable, etc.[/color]

Off topic? Good point, I do apologise. I think I should put a bit more
thought into where I am posting as well as what I am posting.

[Ian Collins]

Re: Stack overflows

Daz wrote:[color=blue]
>
> Off topic? Good point, I do apologise. I think I should put a bit more
> thought into where I am posting as well as what I am posting.
>[/color]
If you are unsure where, mention you platform in your post and someone
will redirect you - but don't make a habit of it :)

--
Ian Collins.

[Steve Pope]

Re: Stack overflows

Phlip <phlipcpp@yahoo .com> wrote:
[color=blue]
>Daz wrote:[/color]
[color=blue][color=green]
>> It is indeed, once again time for me to ask another stupid question.
>>
>> I have been searching around on the net for quite a while, and can't
>> find anything that explains 'exactly' what a stack overflow is, or how
>> it's detected. I know that it's leterally putting too many objects on
>> the stack, but how many objects can you have on the stack, or do OSs
>> only allow a certain number of recursive calls to functions?[/color][/color]
[color=blue]
>You have been posting here long enough to notice all the screams over
>off-topic questions, right?[/color]
[color=blue]
>The limit of your stack is implementation-defined, and the behavior when you
>overflow it is either implementation-defined or undefined.[/color]
[color=blue]
>You could write a program that calls void foo() { foo(); } and run it to see
>what happens. It might be a trappable error, or might be a GPF.[/color]

Somebody correct me if I'm wrong, but I believe the C++ standard
guarantees a recursion depth of 17 -- meaning the in the absence
of stack overflow you can call a function recursively to this
depth, whereas if you exceed the depth it may fail even if
there isn't a stack overflow.

Steve

[Howard]

Re: Stack overflows


"Steve Pope" <spope33@speedy mail.org> wrote in message
news:e7afcm$hci $1@blue.rahul.n et...
[color=blue]
>
> Somebody correct me if I'm wrong, but I believe the C++ standard
> guarantees a recursion depth of 17 -- meaning the in the absence
> of stack overflow you can call a function recursively to this
> depth, whereas if you exceed the depth it may fail even if
> there isn't a stack overflow.
>[/color]

I think you're thinking about the suggested minimum value for the maximum
[guaranteed-to-work] depth of recursive template instantiations, not
anything having to do with the stack or recursive function calls.

-Howard

[Steve Pope]

Re: Stack overflows

Howard <alicebt@hotmai l.com> wrote:
[color=blue]
>"Steve Pope" <spope33@speedy mail.org> wrote in message[/color]
[color=blue][color=green]
>> Somebody correct me if I'm wrong, but I believe the C++ standard
>> guarantees a recursion depth of 17 -- meaning the in the absence
>> of stack overflow you can call a function recursively to this
>> depth, whereas if you exceed the depth it may fail even if
>> there isn't a stack overflow.[/color][/color]
[color=blue]
>I think you're thinking about the suggested minimum value for the maximum
>[guaranteed-to-work] depth of recursive template instantiations, not
>anything having to do with the stack or recursive function calls.[/color]

Thanks, that makes more sense.

Steve