[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: Tor 0.2.1.7-alpha is out



I did a build with Tor 0.2.1.7-alpha about 5 days ago.

Is this the same exact version, or were there updates in the last few days?

Thanks!

- Kyle


On Thu, Nov 20, 2008 at 3:50 PM, Roger Dingledine <arma@xxxxxxx> wrote:
Tor 0.2.1.7-alpha fixes a major security problem in Debian and Ubuntu
packages (and maybe other packages) noticed by Theo de Raadt, fixes
a smaller security flaw that might allow an attacker to access local
services, adds better defense against DNS poisoning attacks on exit
relays, further improves hidden service performance, and fixes a variety
of other issues.

https://www.torproject.org/download.html.en

Changes in version 0.2.1.7-alpha - 2008-11-08
 o Security fixes:
   - The "ClientDNSRejectInternalAddresses" config option wasn't being
     consistently obeyed: if an exit relay refuses a stream because its
     exit policy doesn't allow it, we would remember what IP address
     the relay said the destination address resolves to, even if it's
     an internal IP address. Bugfix on 0.2.0.7-alpha; patch by rovv.
   - The "User" and "Group" config options did not clear the
     supplementary group entries for the Tor process. The "User" option
     is now more robust, and we now set the groups to the specified
     user's primary group. The "Group" option is now ignored. For more
     detailed logging on credential switching, set CREDENTIAL_LOG_LEVEL
     in common/compat.c to LOG_NOTICE or higher. Patch by Jacob Appelbaum
     and Steven Murdoch. Bugfix on 0.0.2pre14. Fixes bug 848.
   - Do not use or believe expired v3 authority certificates. Patch
     from Karsten. Bugfix in 0.2.0.x. Fixes bug 851.

 o Minor features:
   - Now NodeFamily and MyFamily config options allow spaces in
     identity fingerprints, so it's easier to paste them in.
     Suggested by Lucky Green.
   - Implement the 0x20 hack to better resist DNS poisoning: set the
     case on outgoing DNS requests randomly, and reject responses that do
     not match the case correctly. This logic can be disabled with the
     ServerDNSRamdomizeCase setting, if you are using one of the 0.3%
     of servers that do not reliably preserve case in replies. See
     "Increased DNS Forgery Resistance through 0x20-Bit Encoding"
     for more info.
   - Preserve case in replies to DNSPort requests in order to support
     the 0x20 hack for resisting DNS poisoning attacks.

 o Hidden service performance improvements:
   - When the client launches an introduction circuit, retry with a
     new circuit after 30 seconds rather than 60 seconds.
   - Launch a second client-side introduction circuit in parallel
     after a delay of 15 seconds (based on work by Christian Wilms).
   - Hidden services start out building five intro circuits rather
     than three, and when the first three finish they publish a service
     descriptor using those. Now we publish our service descriptor much
     faster after restart.

 o Minor bugfixes:
   - Minor fix in the warning messages when you're having problems
     bootstrapping; also, be more forgiving of bootstrap problems when
     we're still making incremental progress on a given bootstrap phase.
   - When we're choosing an exit node for a circuit, and we have
     no pending streams, choose a good general exit rather than one that
     supports "all the pending streams". Bugfix on 0.1.1.x. Fix by rovv.
   - Send a valid END cell back when a client tries to connect to a
     nonexistent hidden service port. Bugfix on 0.1.2.15. Fixes bug
     840. Patch from rovv.
   - If a broken client asks a non-exit router to connect somewhere,
     do not even do the DNS lookup before rejecting the connection.
     Fixes another case of bug 619. Patch from rovv.
   - Fix another case of assuming, when a specific exit is requested,
     that we know more than the user about what hosts it allows.
     Fixes another case of bug 752. Patch from rovv.
   - Check which hops rendezvous stream cells are associated with to
     prevent possible guess-the-streamid injection attacks from
     intermediate hops. Fixes another case of bug 446. Based on patch
     from rovv.
   - Avoid using a negative right-shift when comparing 32-bit
     addresses. Possible fix for bug 845 and bug 811.
   - Make the assert_circuit_ok() function work correctly on circuits that
     have already been marked for close.
   - Fix read-off-the-end-of-string error in unit tests when decoding
     introduction points.
   - Fix uninitialized size field for memory area allocation: may improve
     memory performance during directory parsing.
   - Treat duplicate certificate fetches as failures, so that we do
     not try to re-fetch an expired certificate over and over and over.
   - Do not say we're fetching a certificate when we'll in fact skip it
     because of a pending download.


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFJJfe061qJaiiYi/URAjQ1AJ9YANIWukD/iWzDf0mhmcdUeFSaywCfa+gh
1Ycg6IFC+DACu48XnQ2nN30=
=64Rm
-----END PGP SIGNATURE-----