[tor-bugs] #7571 [Tor]: Make AutomapHostsOnResolve work with IPv6

Tor Bug Tracker & Wiki blackhole at torproject.org
Mon Dec 17 19:12:39 UTC 2012


#7571: Make AutomapHostsOnResolve work with IPv6
-------------------------+--------------------------------------------------
 Reporter:  nickm        |          Owner:                    
     Type:  enhancement  |         Status:  needs_review      
 Priority:  normal       |      Milestone:  Tor: 0.2.4.x-final
Component:  Tor          |        Version:                    
 Keywords:               |         Parent:                    
   Points:               |   Actualpoints:                    
-------------------------+--------------------------------------------------

Comment(by nickm):

 > Why the loopback by default for IPv4 but link-local for IPv6?

 rransom is right -- there's only one loopback address for IPv6.

 > Other concern: letting 'attempts' be UINT32_MAX for picking random IPv6
 addresses.

 I believe that later in the branch (c50bb9e3), I reduced it to 1000.

 > Is crypto_rand() actually *under our control*?

 The crypto_rand() function uses OpenSSL's cryptographic PRNG.  If that's
 broken (like when Debian broke theirs), or replaced with an LCG or
 something, all OpenSSL users on that platform will be almost totally
 insecure, and an infinite loop in addressmap_get_virtual_address() would
 be the least of our worries ;)

 My inner perfectionist likes the idea of using crit-bit trees, but I'm
 having a hard time constructing a reasonable scenario where 1000 isn't a
 big enough number here.  If that turns out to be wrong, though, we
 shouldn't forget this stuff: maybe we should add a comment or another
 ticket about it.

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/7571#comment:11>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online


More information about the tor-bugs mailing list