Void pointers

Hello,

I am using VC++ and I am testing some C code. I am tying to pass a void pointer !!! I was wondering if someone can help me with the following code sample. I get an "unhandled exception" error at the following line:

P->a = 9;

Here is the full sample code:

Code:

---

#include <stdio.h>

typedef struct tag1{
int a;
}t1;

typedef struct tag2{
int b;
}t2;


void f1(void *i)
{
t1 *P;
P = (t1*)i;        // <<< Type cast!
P->a = 9;          // <<< error here???
}


int main()
{
int x;
t1 *p = NULL;
t2 *q = NULL;

f1(p);
x = p->a;
}

---

Thankyou all in advance for your help!

You are compiling C code as C++ code. In C, void pointers are tremendously useful things. In C++, they are hamstrung by someone's stupid idea of "type safety".

If you compile that code as C, not C++, it should not produce an error.

That isn't the problem. You are passing into the function p, which is NULL. And so you try to dereference a null pointer, which is your problem.

Quote:

---

Originally Posted by MK27

If you compile that code as C, not C++, it should not produce an error.

---

Yeah, instead, it would print "Hello world!" ;)

see the big C: the problem is that p in the main function is a null pointer. You pass it to the f1 function, which then indirectly dereferences it, resulting in undefined behaviour.

EDIT:
Ah, so Elysia beat me to it. No matter, it clearly shows that MK27 is hamstrung by his notion of C++ being hamstrung by an allegedly 'stupid idea of "type safety"' :)

Yup, Laserlight and Elysia... you are both right... a minor modification did the trick:

Code:

---

#include <stdio.h>

typedef struct tag1{
int a;
}t1;

typedef struct tag2{
int b;
}t2;

void f1(void *i)
{
t1 *P;

P = (t1*)i;
P->a = 9;
}

int main()
{
int o;
t1 x;
t2 y;

f1(&x);
o = x.a;
}

---

Which takes me to my next question as to why f1() takes a void* arg. I will start a new thread for that one though.

Thanks everyone for your help!

When you post code in your new thread, please indent it properly.

Quote:

---

Originally Posted by laserlight

No matter, it clearly shows that MK27 is hamstrung by his notion of C++ being hamstrung by an allegedly 'stupid idea of "type safety"' :)

---

Well, it probably did lead me straight to the wrong conclusion there anyway, since this has happened in the past I think with VC++ users.

All apologies.