[tor-bugs] #16467 [Tor]: Faster Ed25519 implementation.

[Tor Bug Tracker & Wiki]

#16467: Faster Ed25519 implementation.
-----------------------------+------------------------------------
     Reporter:  yawning      |      Owner:
         Type:  enhancement  |     Status:  needs_review
     Priority:  normal       |  Milestone:  Tor: 0.2.7.x-final
    Component:  Tor          |    Version:  Tor: 0.2.7
   Resolution:               |   Keywords:  performance, tor-relay
Actual Points:               |  Parent ID:  #9663
       Points:               |
-----------------------------+------------------------------------
Changes (by yawning):

 * status:  new => needs_review


Comment:

 Ok, first pass at: https://github.com/Yawning/tor/compare/feature9663

 It may be sort of annoying to review since the branch also does #9663,
 however only `b61fa4c59c4d3c9226f4514521a5193b77a911dd` really is the
 Curve25519 ntor optimization, the rest falling under this ticket (pulling
 in donna, adding support for our constructs, adding test cases etc).

 I'm not sure if the `struct full of function pointers` approach I took for
 runtime Ed25519 implementation selection was the best way to do things,
 but this is both easy to rip back out, and easy to add another Ed25519
 implementation if someone happens to write `ed25519-super-turbo-2` or
 whatever.

 For Ed25519, the unit tests fuzz all the things, the runtime sanity check
 calls the donna internal test code, and validates that public key
 derivation/signing/checking signatures appears to be working (Maybe this
 should fuzz as well, but I don't think it will really catch anything).

 TODO: Batch verification, leaving it up to nickm (I left a comment in the
 `#if 0`ed out code where there's a bug...)

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/16467#comment:2>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online