I am confused with pointer type casting. I read that i cannot type case an integer pointer to float pointer. Am I correct? The code is
or can i take blindly that pointers should never be type casted?Code:#include <stdio.h> int main(void) { int value=0x20FE; float *fp; int *p = &value; fp = (float *)(&value); printf("int value = %f\n", fp); printf("char ptr value = %d\n",p); return 0; }
The relevant rule is:Originally Posted by Satya
So, if the alignment works out, you can cast a pointer to int to pointer to float. But there would be nothing else useful you could do with it other than converting it back again since the representation of an int likely makes no sense when reinterpreted as a float, and that's what you would be doing with the cast: reinterpreting the int that the pointer points to as if it were a float.
No, as you can see from the rule, you can safely cast your pointer to int to be a pointer to char (or unsigned char, or signed char). This makes sense as you're reinterpreting the int as a sequence of bytes... but that is exactly what the underlying storage of the int consists of! Additionally:
Look up a C++ Reference and learn How To Ask Questions The Smart Way
Thank you very much for the reply. Can you please explain me
with one example? How do i know whether it is aligned or not aligned correctly? Please help."If the resulting pointer is not correctly aligned for the referenced type,"
Consider:
Assuming that float is 4 bytes and double is 8 bytes, p now points to an object that isn't actually big enough for p to point to.Code:float value = 1.23f; double *p = (double*)(&value);
Look up a C++ Reference and learn How To Ask Questions The Smart Way
