Hi,
On Thu, Oct 08, 2015 at 03:50:46PM +0200, David Herrmann wrote:
> Hi
>
> On Thu, Oct 8, 2015 at 1:31 PM, Sergei Zviagintsev <ser...@s15v.net> wrote:
> > Add comment on why we remove the same slice from free slices tree and
> > then add it back again when merging the slice to be released with
> > previous free slice.
> >
> > Signed-off-by: Sergei Zviagintsev <ser...@s15v.net>
> > ---
> > ipc/kdbus/pool.c | 6 ++++++
> > 1 file changed, 6 insertions(+)
> >
> > diff --git a/ipc/kdbus/pool.c b/ipc/kdbus/pool.c
> > index 84afe96fbc22..c26ef963d8d1 100644
> > --- a/ipc/kdbus/pool.c
> > +++ b/ipc/kdbus/pool.c
> > @@ -304,6 +304,12 @@ static void __kdbus_pool_slice_release(struct
> > kdbus_pool_slice *slice)
> > s = list_entry(slice->entry.prev,
> > struct kdbus_pool_slice, entry);
> > if (s->free) {
> > + /*
> > + * As size of slice increases after merge and free
> > + * slices tree is ordered by slice size, we have to
> > + * remove the slice from free slices tree and then
> > add
> > + * it again to keep the tree balanced.
> > + */
>
> Mhh, isn't this obvious? "slices_free" is ordered by s->size, so a
> change of the key requires a re-insert. If you disagree, maybe keep it
> simple:
>
> /* s->size changed, re-insert slice in rbtree */
It wasn't obvious for me from the first sight (but don't tell anyone),
so I decided that some words on why we're doing re-insert would be
helpful. It's up to you to decide whether it would be useful for the
others. BTW, I find your shorter variant better.
>
> Thanks
> David
>
> > rb_erase(&s->rb_node, &pool->slices_free);
> > list_del(&slice->entry);
> > s->size += slice->size;
> > --
> > 1.8.3.1
> >
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
