[TUHS] C++ / Kernel

[Bakul Shah]

On Aug 24, 2018, at 5:17 AM, ron at ronnatalie.com wrote:
> 
> For instance:
>       int x;
>       daddr_t* y;
> 
>      u.b_un.b_words = &x;
>      y = u.b_un.b_daddr;

One would typically *not* do this.

> 
> This is syntactically correct, but the code will blow up bizarrely on
> machines where the int* and long* pointers are not in the same format. 
> Now if the datatype was a void* and we did this:
> 
>      u.b_vp = & x;
>      y = u.b_vp;
> 
> Provided there weren't any alignment issues, this would work as the
> conversion to and from void* would result in the right pointer value.

For example, given

	union {
	    caddr_t b_addr;		/* low order core address */
	    int	*b_words;		/* words for clearing */
	    struct filsys *b_filsys;	/* superblocks */
	    struct dinode *b_dino;	/* ilist */
	    daddr_t *b_daddr;		/* indirect block */
	} b_un;

You'd use it something like this:

	struct b_un x, y, z, *w;
	struct filesys fs;
	daddr_t d;
	...
	x.b_words = &b;
	y.b_filsys = &fs;
	z.b_addr = &d;
	w = &x;

These come about because in different contexts different ptrs
may be used but one would not do any type conversion. Said another
way, if you replaced the union with a struct, type punning code
would fall apart but legit code would continue working (except
in the case where the *size* of the object is significant).