Dynamic hash table size (with static has load)

[Wayne Davison]

On Sun, Feb 26, 2006 at 06:55:18PM +0200, Shachar Shemesh wrote:
> - Attacker puts just enough checksums in the count to make the 20%
> expansion overflow the 32bit variable.

Quite so.  I also remembered the 20% expansion after I sent my email,
which (as you mentioned) does mean that we need to bump the variable up
to at least a uint32.  Thanks for your persistence in making sure that I
get this right.

This would be such an improvement for really large files that I'm
considering putting this into 2.6.7pre3.  However, I want to be sure
that it is in rock-solid shape first.  If anyone wants to help with the
testing and/or the discussion of its acceptance into 2.6.7, please feel
free to dig in and send us your comments on the list.  Thanks,

..wayne..