[tor-bugs] #4900 [Tor]: Use a more randomized hash function for our hash tables

[Tor Bug Tracker & Wiki]

#4900: Use a more randomized hash function for our hash tables
-----------------------------+------------------------------------
     Reporter:  nickm        |      Owner:
         Type:  enhancement  |     Status:  new
     Priority:  major        |  Milestone:  Tor: 0.2.5.x-final
    Component:  Tor          |    Version:
   Resolution:               |   Keywords:  dos security tor-relay
Actual Points:               |  Parent ID:
       Points:               |
-----------------------------+------------------------------------

Comment (by nickm):

 bug4900_siphash in my public repository has an incomplete siphash-based
 implementation for initial poking-at.  Initially, it looks a little
 costly, but I didn't use a screamingly optimized implementation or
 anything.  It doesn't randomize its keys yet, so it's not ready for prime-
 time.

 Possibly, we should also consider crit-bit trees, though those have
 potential timing issues that we would need to consider hard.  And whatever
 else about them I forgot.  I wonder where I put my incomplete critbit-
 based HT_* replacement code, and whether it's worth finishing.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/4900#comment:11>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online