[PATCH] Add ffs() to libiberty

[DJ Delorie]

> In the _subsequent_ tests.  The (valu & 0xCC) and (valu & 0xAA)
> tests aren't valid unless there is exactly one bit set.

Doh, you're right.

> On the contrary, it's well defined.

Still, I'd rather see something like this:

	valu = valu & ~(valu - 1);

than see negation of an unsigned value (yes, I know they *do* exactly
the same thing, but they *mean* different things).

We probably should put a comment on that line too, like "Make sure
there is exactly one bit set."