Exactly:
Accessing MY_8255_BASE[0] accesses the Port-A Register of the 8255;
Accessing MY_8255_BASE[1] accesses the Port-B Register;
Accessing MY_8255_BASE[2] accesses the Port-C Register;
Accessing MY_8255_BASE[3] accesses the Control Register.

The array alone gives you access to all the registers!

So what is the point of adding a pointer:

unsigned char *ptr_8255;

void init_8255(unsigned char control_word)
{
	ptr_8255 = MY_8255_BASE;
	*(ptr_8255 + 3) = control_word;
}
 

It just adds an unnecessary level of indirection - to no benefit!

Wouldn't it be far simpler & clearer to just write:

void init_8255(unsigned char control_word)
{
        MY_8255_BASE[3] = control_word;

}
 

You could also make it clearer by defining symbolic names for the array indexes; eg,

enum
{
   PPI_PORT_A  = 0, // Index for Port-A
   PPI_PORT_B  = 1, // Index for Port-B
   PPI_PORT_C  = 2, // Index for Port-C
   PPI_CONTROL = 3  // Index for the Control register
}

 

And then have:

void init_8255(unsigned char control_word)
{
	MY_8255_BASE[ PPI_CONTROL ]  = control_word;
}
 

Which doesn't even really need to be in a function call...