Pointers - *p++

Hi. I'm trying to understand what my professor says when he refers to
declaring a pointer, and seeing the operator precedence and seeing
what happens when we do things like

p++
*p++
(*p)++ etc,

but he starts out with this declaration
int *p = (int *)0;

Although my compiler does not complain, when I try to do something
like
cout << *p;

I get a Bus Error on the console (I'm using Xcode). What am I doing
wrong here? Thanks.

crystal twix

crystal twix wrote:
> Hi. I'm trying to understand what my professor says when he refers to
> declaring a pointer, and seeing the operator precedence and seeing
> what happens when we do things like
>
> p++
> *p++
> (*p)++ etc,
>
> but he starts out with this declaration
> int *p = (int *)0;
>
> Although my compiler does not complain, when I try to do something
> like
> cout << *p;
>
> I get a Bus Error on the console (I'm using Xcode). What am I doing
> wrong here? Thanks.

Dereferencing a NULL pointer.

red floyd

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

crystal twix wrote:

> Hi. I'm trying to understand what my professor says when he refers to
> declaring a pointer, and seeing the operator precedence and seeing
> what happens when we do things like
>
> p++
> *p++
> (*p)++ etc,
>
> but he starts out with this declaration
> int *p = (int *)0;
Better to write "int *p=NULL;" instead
>
> Although my compiler does not complain, when I try to do something
> like
> cout << *p;
p is a pointer object. As the value of p is NULL, dereferencing it becomes
an object that does not exist. The output operation tries to take the value
in the object, however, as the object does not exist, the behaviour is
undefined.
>
> I get a Bus Error on the console (I'm using Xcode). What am I doing
> wrong here? Thanks.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkru724ACgkQG6NzcAXitM+4kgCgkS2jciV057 fHKMZoMzVT6iNK
yjgAnRRhKLN/i+EOEZOX87I6V947BiR9
=V0pC
-----END PGP SIGNATURE-----

Michael Tsang

> Better to write "int *p=NULL;" instead

I am *not* saying anything against but, I write int* p=0; as it remind
me that in C++ 'NULL' is 'zer0'.

siddhant3s

On Nov 2, 6:44*pm, siddhant3s <siddhan...@gmail.com> wrote:
> > Better to write "int *p=NULL;" instead
>
> I am *not* saying anything against but, I write int* p=0; as it remind
> me that in C++ 'NULL' is 'zer0'.

This is not always true. It depends of the implementation of your C++.

paperab

crystal twix wrote:
> Hi. I'm trying to understand what my professor says when he refers to
> declaring a pointer, and seeing the operator precedence and seeing
> what happens when we do things like
>
> p++
> *p++
> (*p)++ etc,
>
> but he starts out with this declaration
> int *p = (int *)0;

If you get into the habit of declaring and initialising pointers at the
point of first use, the discussion is moot.

--
Ian Collins

Ian Collins

On Nov 3, 12:38 am, paperab <ab9380...@yahoo.co.uk> wrote:
> On Nov 2, 6:44 pm, siddhant3s <siddhan...@gmail.com> wrote:

> > > Better to write "int *p=NULL;" instead

> > I am *not* saying anything against but, I write int* p=0; as
> > it remind me that in C++ 'NULL' is 'zer0'.

Yes. It's a style issue. I too prefer NULL, but I understand
people who prefer 0. Either is "idiomatic". The original:
int *p = (int*)0;
isn't.

> This is not always true. It depends of the implementation of
> your C++.

It's not necessarily true that NULL expands to the exact string
0, but it must expand to a null pointer constant, i.e. an
integral constant expression evaluating to 0. (It's interesting
that a null pointer constant is not allowed to be a pointer.) 0
is by far the most frequent solution, and 0L was also prevelant
at one time. At least one compiler uses something like
__builtin_null, so that it can warn if NULL is used as an
integer; this is by far the best solution.

--
James Kanze

James Kanze

On Nov 3, 3:50 am, Ian Collins <ian-n...@hotmail.com> wrote:
> crystal twix wrote:
> > Hi. I'm trying to understand what my professor says when he
> > refers to declaring a pointer, and seeing the operator
> > precedence and seeing what happens when we do things like

> > p++
> > *p++
> > (*p)++ etc,

> > but he starts out with this declaration
> > int *p = (int *)0;

> If you get into the habit of declaring and initialising
> pointers at the point of first use, the discussion is moot.

He did. He initialized it with a null pointer, then proceeded
with a number of operations which aren't legal on null pointers.

--
James Kanze

James Kanze

On Nov 3, 5:38*am, paperab <ab9380...@yahoo.co.uk> wrote:
> On Nov 2, 6:44*pm, siddhant3s <siddhan...@gmail.com> wrote:
>
> > > Better to write "int *p=NULL;" instead
>
> > I am *not* saying anything against but, I write int* p=0; as it remind
> > me that in C++ 'NULL' is 'zer0'.
>
> This is not always true. It depends of the implementation of your C++.

Read http://www.research.att.com/~bs/bs_faq2.html#null

siddhant3s