Re: [tor-bugs] #9699 [Tor]: tor curve25519 fail

["Tor Bug Tracker & Wiki" <blackhole@xxxxxxxxxxxxxx>]

#9699: tor curve25519 fail
-----------------------------+------------------------------
     Reporter:  cypherpunks  |      Owner:
         Type:  defect       |     Status:  new
     Priority:  normal       |  Milestone:  Tor: unspecified
    Component:  Tor          |    Version:
   Resolution:               |   Keywords:  tor-client
Actual Points:               |  Parent ID:
       Points:               |
-----------------------------+------------------------------

Comment (by tmpname0901):

 Replying to [comment:4 nickm]:
 > tmpname0901: Can you think of any simple check to detect the broken
 default centos 5.9 compiler?  If you can find a tiny program that the
 default compiler refuses to compile correctly, we could use that to detect
 a broken implementation at build time from the configure script.

 No, I don't have a test case, beyond the obvious whittling down of
 curve25519-donna-c64.c until the mis-compilation is no longer seen.  Or
 maybe playing with the compiler switches.

 I use gcc44 not because I fear the buggy default compiler but because the
 newer version of GCC has better support for contemporary CPUs and general
 optimization.  (GCC v4.1 was released in Feb. 2007 - that's a long time
 ago in CPU generations.)  Tor can be CPU-intensive and I want to give it
 every advantage that the hardware can provide.

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/9699#comment:5>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs