On Thu, Jul 12, 2012 at 7:12 AM, Peter Zijlstra [off-list ref] wrote:

On Wed, 2012-07-11 at 18:12 -0700, Michel Lespinasse wrote:

quoted

In __rb_erase_color(), some of the cases are more complicated than you drew however, because some node colors aren't known.

Right, the wikipedia article draws them blank, I couldn't come up with a
3rd case, although maybe we can annotate them like (P) to mean blank..

Ah, good idea, I adopted that :)

Yes, very nice.. someday when I'm bored I might expand the comments with
the reason why we're doing the given operation.

There is a brief comment at the start of the loop that indicates which
rbtree invariants might be violated at that point; so someone could
deduce that we're trying to either fix these or move towards the root
until they get fixed. But yeah, this is never explicitly explained.

Also, I was sorely tempted to rename your tmp1,tmp2 variables to sl and
sr.

This could be done, but you'd *still* need one extra temporary, so
you'd end up with sl, sr and tmp. Which is fine, I guess, but I
preferred to have one less variable around.

-- 
Michel "Walken" Lespinasse
A program is never fully debugged until the last user dies.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>