On Wed, Feb 23, 2022 at 12:19 PM Rasmus Villemoes
<linux@xxxxxxxxxxxxxxxxxx> wrote:
>
> I have often wished that the iterator macros would consistently set the
> loop variable to NULL upon reaching the end.
I really think the rule should be that to a 99% approximation, we
should strive only ever use the iterated-upon value *inside* the loop.
No, that's now how we do it now. But I think the "break out and do the
work outside the loop" case is kind of broken anyway. It makes you
test the condition twice - and while a compiler might be smart enough
to optimize the second test away, it's still just plain ugly.
Linus
