> Research work shows that only at around 1000 entires does the lookup
> speed make up for the balancing cost of a tree.
Was this research done for the Linux VMA case or for some "average" access
patterns used by the database folks? The latter would be of little use since
in most cases we have more finds than inserts/deletes (*much* more on a i386
where we need it for access_ok()).
Two of the sample cases (amongst others) were Linux VMA and Schoebel's
new I/O layer subsystem. The others were scattered over various other
applications such as routing table lookups etc.
Later,
David S. Miller
davem@dm.cobaltmicro.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/