Re: Avoiding dangling pointers.

Rafael Anschau wrote:

> I read that you should assign null (0) to all pointers that you call
> delete on.
> 
> Does that mean:
> 
> *p=0(set the value pointed to to 0).
> 
> or
> 
> p=0(set the address held to zero).

This one.

> The last one is awkward for none would want a pointer pointing
> to address zero(even though it compiles fine, you can´t assign
> anything to it).

The rationale is that a null pointer is much easier to debug than a
dangling
pointer. When you access a null pointer, on most platforms it send the
program to death (or debugger), but using a dangling pointer can overwrite
someone else's data which can manifest itself much much later. You should
certainly do it if you know you're gonna read the pointer later.

But don't take the advice (about setting pointers to zero) too seriously.
Basically, if your program has the memory management stuff correct (which
it really, really should), it is just superfluous. If you've deleted an
object through one of the ten pointers to it, the other nine are still
left
dangling. Also, if you use valgrind to debug your code, it will show you
detailed info about accesses through dangling pointers (like backtraces
from the time it was allocated and freed), but not if you set it to 0.

It's like uninitialised variables: you can always get rid of uninitialised
variable accesses by initialising every variable to 0, but it's up to you
to decide for a particular variable whether it makes sense or it's just
hiding a bug.

Regards
     Jiri Palecek


-- 
      [ See http://www.gotw.ca/resources/clcm.htm
for info about ]
      [ comp.lang.c++.moderated.    First time posters: Do this! ]