Re: [RFC] Generalized prio_tree, revisited

[Werner Almesberger <werner@almesberger.net>]

Rajesh Venkatasubramanian wrote:
> I wonder whether we should use [start, last]

Yes, good idea. I've changed it in my tree.

> prio_tree_replace should be static in prio_tree.c.

Indeed. Thanks !

> > +struct prio_tree_node *prio_tree_first(struct prio_tree_iter *iter);
> 
> Should we go with prio_tree_iter_init and remove prio_tree_first
> (similar to vma_prio_tree_next) ? I am not very particular about it,
> though.

You mean to roll prio_tree_first and prio_tree_iter_init into a
single call, so that prio_tree_first would look similar to the
one in 2.6.7 ?

> > +static void get_index(const struct prio_tree_root *root,
> 
> Should be "inline" ?

That's of course what we hope to happen, but I'd leave the inlining
decision to the compiler. After all, it's supposed to be really
good at such things nowadays ;-)

Thanks,
- Werner

-- 
  _________________________________________________________________________
 / Werner Almesberger, Buenos Aires, Argentina     werner@almesberger.net /
/_http://www.almesberger.net/____________________________________________/