intel: Massively simplify the BO cache.HEADmaster

At least since I added the sizes in between the powers of two, the
whole bucketing idea to try and approach O(1) search in the cache has
been defeated by an up-to-56-step search for which bucket to look in.

Instead, just walk up to the whole cache every time.  We don't expect
the number of entries in the cache to be that large, and it means we
don't need to round up our initial allocation size, saving memory.

0 files changed, 0 insertions, 0 deletions