]>
git.ipfire.org Git - thirdparty/tor.git/log
Mike Perry [Mon, 10 Dec 2012 07:58:01 +0000 (23:58 -0800)]
Update changes file.
Mike Perry [Mon, 10 Dec 2012 07:50:05 +0000 (23:50 -0800)]
Clean up some XXX comments.
Mike Perry [Mon, 10 Dec 2012 07:47:04 +0000 (23:47 -0800)]
Space fixes.
Mike Perry [Mon, 10 Dec 2012 04:56:48 +0000 (20:56 -0800)]
Fix a rather serious use-count state bug.
We need to use the success count or the use count depending on the consensus
parameter.
Mike Perry [Mon, 10 Dec 2012 04:53:22 +0000 (20:53 -0800)]
Convert to doubles for all pathbias state.
Let's hope this solves the rounding error issue..
Mike Perry [Mon, 10 Dec 2012 04:20:44 +0000 (20:20 -0800)]
Also shorten circuit_successes to circ_successes.
For consistency and great justice.
Ok, mostly consistency.
Mike Perry [Mon, 10 Dec 2012 04:18:31 +0000 (20:18 -0800)]
Rename first_hop to circ_attempt.
Since we've generalized what we can count from (first or second hop), we
should generalize the variable and constant naming too.
Mike Perry [Mon, 10 Dec 2012 04:02:41 +0000 (20:02 -0800)]
Change from first hop accounting to 2nd hop accounting
This has several advantages, including more resilience to ambient failure.
I still need to rename all the first_hop vars tho.. Saving that for a separate
commit.
Mike Perry [Mon, 10 Dec 2012 04:02:08 +0000 (20:02 -0800)]
Add intro+rend cannibalize param..
Mike Perry [Mon, 10 Dec 2012 03:18:04 +0000 (19:18 -0800)]
Flag cannibalized circs as used (non-ideal).
Also add some comments.
Mike Perry [Sun, 9 Dec 2012 00:37:22 +0000 (16:37 -0800)]
Allow any valid 'end' cell to mean a circuit was used successfully.
Also improve some log messages.
Mike Perry [Sat, 8 Dec 2012 22:16:29 +0000 (14:16 -0800)]
Sadly, we can't safely count client intro circ success
Mike Perry [Sat, 8 Dec 2012 20:07:58 +0000 (12:07 -0800)]
Fix some hidden service edge cases.
Mike Perry [Sat, 8 Dec 2012 01:47:23 +0000 (17:47 -0800)]
Refactor path use bias code into own function.
Also, improve and log some failure cases.
Mike Perry [Fri, 7 Dec 2012 23:50:31 +0000 (15:50 -0800)]
Actually, both nacks and acks indicate a valid path
Mike Perry [Mon, 26 Nov 2012 01:29:16 +0000 (17:29 -0800)]
Note more potential issues.
Mike Perry [Thu, 22 Nov 2012 00:40:25 +0000 (16:40 -0800)]
Note a strange case for SOCKS streams.
Mike Perry [Thu, 22 Nov 2012 00:33:16 +0000 (16:33 -0800)]
Fix another crash bug.
Mike Perry [Thu, 22 Nov 2012 00:32:38 +0000 (16:32 -0800)]
Fix a crash bug and pass down a remote reason code.
Unexpected channel closures count as remote circ failures.
Mike Perry [Thu, 22 Nov 2012 00:31:58 +0000 (16:31 -0800)]
Document that care needs to be taken with any_streams_attached.
Mike Perry [Tue, 20 Nov 2012 09:52:33 +0000 (01:52 -0800)]
Remove n_chan codepaths for determinining guard.
Cpath is apparently good enough.
Mike Perry [Mon, 19 Nov 2012 19:31:35 +0000 (11:31 -0800)]
Move a pathbias function that depends on entryguard_t.
Mike Perry [Mon, 19 Nov 2012 19:30:07 +0000 (11:30 -0800)]
Fix a crash due to NULL circ->n_chan.
Is this redundant? Can we always rely on circ->cpath->extend_info
being present for origin circuits?
Mike Perry [Mon, 19 Nov 2012 18:45:47 +0000 (10:45 -0800)]
Prop209: Rend circuits weren't ever marked dirty.
Mike Perry [Sun, 18 Nov 2012 01:51:27 +0000 (17:51 -0800)]
Prop 209: Add in hidserv path bias counts for usage.
Mike Perry [Sun, 18 Nov 2012 00:30:50 +0000 (16:30 -0800)]
Prop 209: Add path bias counts for timeouts and other mechanisms.
Turns out there's more than one way to block a tagged circuit.
This seems to successfully handle all of the normal exit circuits. Hidden
services need additional tweaks, still.
Mike Perry [Fri, 2 Nov 2012 19:37:26 +0000 (12:37 -0700)]
Be explicit about units for timeout.
Mike Perry [Fri, 2 Nov 2012 19:36:08 +0000 (12:36 -0700)]
Fix an assert crash and an incorrectly placed return.
Mike Perry [Thu, 1 Nov 2012 01:51:07 +0000 (18:51 -0700)]
Add a changes file for bug7157.
Mike Perry [Thu, 1 Nov 2012 01:50:45 +0000 (18:50 -0700)]
Update manpage for new PathBias torrc options.
Mike Perry [Thu, 1 Nov 2012 01:49:49 +0000 (18:49 -0700)]
Update with code review changes from Nick.
Mike Perry [Thu, 25 Oct 2012 21:14:28 +0000 (14:14 -0700)]
Fix spaces.
Mike Perry [Thu, 25 Oct 2012 21:05:44 +0000 (14:05 -0700)]
Refactor pathbias functions to use pathbias_should_count.
Mike Perry [Thu, 25 Oct 2012 20:42:37 +0000 (13:42 -0700)]
Update Path Bias log messages to match Proposal 209.
Mike Perry [Thu, 25 Oct 2012 01:15:41 +0000 (18:15 -0700)]
Add log message checks for different rates.
May want to squash this forward or back..
Mike Perry [Thu, 25 Oct 2012 01:03:09 +0000 (18:03 -0700)]
Update pathbias parameters to match Proposal 209.
Needs manpage update and testing still..
Mike Perry [Thu, 25 Oct 2012 00:34:18 +0000 (17:34 -0700)]
Add the ability to count circuit timeouts for guards.
This is purely for informational reasons for debugging.
Nick Mathewson [Fri, 7 Dec 2012 19:14:20 +0000 (14:14 -0500)]
Fix some wide lines
Nick Mathewson [Fri, 7 Dec 2012 19:12:17 +0000 (14:12 -0500)]
Merge branch 'win64-7260'
Conflicts:
src/or/dns.c
Nick Mathewson [Fri, 7 Dec 2012 19:08:07 +0000 (14:08 -0500)]
Fix infinite loop in circuit_expire_bulding
Fixes bug 7663; bug introduced in
42e3c04a7a5fb47a9 . Not in any
released version of Tor.
Nick Mathewson [Fri, 7 Dec 2012 17:42:02 +0000 (12:42 -0500)]
Expand the manpage to say: watch out for nonlocal socks
This clears up the remaining issue stopping me from closing bug 6297.
Nick Mathewson [Fri, 7 Dec 2012 16:39:56 +0000 (11:39 -0500)]
Merge remote-tracking branch 'karsten/task-6266'
Nick Mathewson [Fri, 7 Dec 2012 16:11:26 +0000 (11:11 -0500)]
Merge branch 'bug7306'
Nick Mathewson [Fri, 7 Dec 2012 16:02:27 +0000 (11:02 -0500)]
Merge remote-tracking branch 'public/bug6887'
Nick Mathewson [Fri, 7 Dec 2012 15:35:41 +0000 (10:35 -0500)]
Merge branch 'bug3443_squashed'
Mike Perry [Fri, 26 Oct 2012 00:43:10 +0000 (17:43 -0700)]
Bug 3443: Don't count ORconn setup in circuit build time.
Also, add a hack Roger suggested where we're more patient if no circuits are
opened yet.
Nick Mathewson [Thu, 6 Dec 2012 15:59:02 +0000 (10:59 -0500)]
Use FreeLibrary, not CloseHandle, for library in test_util.c
Fix for bug 7306. Bugfix on 0.2.2.17-alpha.
Karsten Loesing [Thu, 6 Dec 2012 10:29:47 +0000 (11:29 +0100)]
Update to the December 2012 GeoIP database.
Nick Mathewson [Wed, 5 Dec 2012 02:47:45 +0000 (21:47 -0500)]
Merge remote-tracking branch 'asn/bug7592_take2'
George Kadianakis [Wed, 5 Dec 2012 02:32:11 +0000 (04:32 +0200)]
Return connection_exit_connect() if payload creation failed.
Fixes bug #7592; bugfix on
882b389668067a29bb539d0f5bd5cb2f83b93012 .
The bug is not present in any released versions of Tor.
Roger Dingledine [Mon, 3 Dec 2012 18:33:43 +0000 (13:33 -0500)]
fix some typos
Nick Mathewson [Wed, 28 Nov 2012 03:18:16 +0000 (22:18 -0500)]
Merge branch 'bug7013_take2_squashed'
George Kadianakis [Tue, 27 Nov 2012 22:24:58 +0000 (00:24 +0200)]
Introduce tor_addr_port_parse() and use it to parse ServerTransportListenAddr.
George Kadianakis [Tue, 30 Oct 2012 02:17:13 +0000 (04:17 +0200)]
Add a torrc option to specify the bind address of managed proxies.
Karsten Loesing [Wed, 28 Nov 2012 02:26:52 +0000 (21:26 -0500)]
Update to the November 2012 GeoIP database.
Karsten Loesing [Wed, 28 Nov 2012 02:22:58 +0000 (21:22 -0500)]
Add script to fix "A1" entries in geoip file.
Fixes #6266.
Nick Mathewson [Fri, 23 Nov 2012 16:36:44 +0000 (11:36 -0500)]
Make sure that the error in ADDRMAP events is well-formed
"error=Unable to launch resolve request" is not a nice thing to tell
the controller. Bugfix on 0.2.0.19-alpha (
c11c48fc ).
Nick Mathewson [Fri, 23 Nov 2012 15:51:11 +0000 (10:51 -0500)]
Minor documentation fix
Nick Mathewson [Fri, 23 Nov 2012 15:06:16 +0000 (10:06 -0500)]
Note limitation of parse_rfc_1123_time
RFC1123 suggests that we should handle two-year times, and a full
range of time zones, and other stuff too. We don't.
Nick Mathewson [Fri, 23 Nov 2012 15:05:16 +0000 (10:05 -0500)]
In comments and logs, say "UTC" not "GMT"
Fix for #6113.
Note that the RFC1123 times we generate still all say 'GMT'. I'm
going to suggest this is not worth changing.
Nick Mathewson [Fri, 23 Nov 2012 14:51:31 +0000 (09:51 -0500)]
Refer to RFC 4648 instead of the obsolete RFC 3548
Affects comments only. For ticket 6849.
Roger Dingledine [Tue, 20 Nov 2012 08:46:56 +0000 (03:46 -0500)]
forward-port the 0.2.3.25 changelog and release notes
Nick Mathewson [Sun, 18 Nov 2012 23:46:57 +0000 (18:46 -0500)]
Merge branch 'bug7493_redux'
Nick Mathewson [Sun, 18 Nov 2012 22:15:41 +0000 (17:15 -0500)]
Initialize ipv{4,6}_traffic_ok in entry_connection_new
This one is necessary for sending BEGIN cells with sane flags when
self-testing a directory port. All real entry connections were
getting their ipv{4,6}_traffic_ok flags set from their listeners, and
for begindir entry connections we didn't care, but for directory
self-testing, we had a problem.
Fixes at least one more case of 7493; if there are more lingering
cases of 7493, this might fix them too.
Bug not in any released version of Tor.
Roger Dingledine [Fri, 16 Nov 2012 16:38:56 +0000 (11:38 -0500)]
when counting available descs, say whether we're counting exits
Nick Mathewson [Fri, 16 Nov 2012 03:58:54 +0000 (22:58 -0500)]
Give useful warning when both IPv4 and IPv6 are disabled on a socksport
Nick Mathewson [Fri, 16 Nov 2012 03:49:43 +0000 (22:49 -0500)]
Allow IPv4 traffic on default and old-style-config SocksPorts.
Looks like when i was writing the code to set the ipv4_traffic flag on
port_cfg_t, I missed some cases, such as the one where the port was
set from its default value.
Fix for 7493. Bug not in any released Tor.
Nick Mathewson [Thu, 15 Nov 2012 19:45:15 +0000 (14:45 -0500)]
Merge branch 'ipv6_exits'
Nick Mathewson [Thu, 15 Nov 2012 18:00:19 +0000 (13:00 -0500)]
Set IPv4/IPv6 flags correctly when being a SOCKS client
Nick Mathewson [Thu, 15 Nov 2012 17:17:36 +0000 (12:17 -0500)]
Remove some XXXX commens in dns.c
Previously, I was freaking out about passing an unspec address to
dns_found_answer() on an error, since I was using the address type to
determine whether the error was an error on an ipv4 address lookup or
on an ipv6 address lookup. But now dns_found_answer() has a separate
orig_query_type argument to tell what kind of query it is, so there's
no need to freak out.
Nick Mathewson [Thu, 15 Nov 2012 17:17:30 +0000 (12:17 -0500)]
Fix up some comments in connection_edge.c
Nick Mathewson [Thu, 15 Nov 2012 04:35:55 +0000 (23:35 -0500)]
Add a changes file for IPv6 exits
Nick Mathewson [Thu, 15 Nov 2012 04:35:13 +0000 (23:35 -0500)]
Add manual page entries for new IPv6-exits options
Nick Mathewson [Thu, 15 Nov 2012 03:55:23 +0000 (22:55 -0500)]
Add another missing function doc
Nick Mathewson [Thu, 15 Nov 2012 03:29:08 +0000 (22:29 -0500)]
Initial support for AAAA requests on DNSPort.
This is imperfect, since it sends back whatever we would send to
a socks RESOLVE request, when in reality we should send back whatever
was asked for.
Nick Mathewson [Thu, 15 Nov 2012 03:20:18 +0000 (22:20 -0500)]
Let tor-resolve generate PTR requests for IPv6 addresses
Nick Mathewson [Thu, 15 Nov 2012 03:20:01 +0000 (22:20 -0500)]
Accept reverse resolve requests for IPv6 addresses
Nick Mathewson [Thu, 15 Nov 2012 03:07:18 +0000 (22:07 -0500)]
Add IPv6 support to tor-resolve so it can hear about IPv6 answers
Nick Mathewson [Thu, 15 Nov 2012 03:06:13 +0000 (22:06 -0500)]
Make DNS resolve requests work for IPv6
* If there's an IPv4 and an IPv6 address, return both in the resolved
cell.
* Treat all resolve requests as permitting IPv6, since by the spec they're
allowed to, and by the code that won't break anything.
Nick Mathewson [Thu, 15 Nov 2012 01:51:41 +0000 (20:51 -0500)]
Fix a bug in policy_is_reject_star() that was making IPv4 exits break
IPv4-only exits have an implicit "reject [::]/0", which was making
policy_is_reject_star() return 1 for them, making us refuse to do
hostname lookups.
This fix chanes policy_is_reject_star() to ask about which family we meant.
Nick Mathewson [Wed, 14 Nov 2012 20:20:28 +0000 (15:20 -0500)]
Add some missing doxygen for ipv6 exit code
Nick Mathewson [Wed, 14 Nov 2012 15:48:58 +0000 (10:48 -0500)]
Actually send back correctly-formed IPv6 CONNECTED cells
We had some old code to send back connected cells for IPv6 addresses,
but it was wrong. Fortunately, it was also unreachable.
Nick Mathewson [Wed, 14 Nov 2012 15:18:33 +0000 (10:18 -0500)]
Remove some unused defines in dns.c
Nick Mathewson [Wed, 14 Nov 2012 15:14:23 +0000 (10:14 -0500)]
Repair DNS NEXIST hijacking workaround
The code previously detected wildcarding and replaced wildcarded
answers with DNS_STATUS_FAILED_PERMANENT. But that status variable
was no longer used! Remove the status variable, and instead change
the value of 'result' in evdns_callback.
Thank goodness for compiler warnings. In this case,
unused-but-set-variable.
Thanks to Linus for finding this one.
Nick Mathewson [Wed, 14 Nov 2012 15:09:06 +0000 (10:09 -0500)]
Implement a PreferIPv6 flag for SocksPorts
Nick Mathewson [Tue, 6 Nov 2012 19:30:30 +0000 (14:30 -0500)]
Define a wrapper for evdns_base_resolve_ipv6 for systems w/o libevent 2
Nick Mathewson [Tue, 6 Nov 2012 18:35:31 +0000 (13:35 -0500)]
Make address_is_invalid_destination recognize ipv6 addrs as valid.
Nick Mathewson [Tue, 6 Nov 2012 17:12:59 +0000 (12:12 -0500)]
Actually generate microdescriptors with p6 lines.
Nick Mathewson [Tue, 6 Nov 2012 17:12:47 +0000 (12:12 -0500)]
Remove a since-fixed XXX; improve a doxygen comment
Nick Mathewson [Tue, 6 Nov 2012 16:29:41 +0000 (11:29 -0500)]
Actually advertise IPv6 exit policies.
I have a theory that my tests will work better if the code I'm testing
isn't disabled.
Nick Mathewson [Mon, 5 Nov 2012 18:26:29 +0000 (13:26 -0500)]
Revise the DNS subsystem to handle IPv6 exits.
Now, every cached_resolve_t can remember an IPv4 result *and* an IPv6
result. As a light protection against timing-based distinguishers for
IPv6 users (and against complexity!), every forward request generates
an IPv4 *and* an IPv6 request, assuming that we're an IPv6 exit. Once
we have answers or errors for both, we act accordingly.
This patch additionally makes some useful refactorings in the dns.c
code, though there is quite a bit more of useful refactoring that could
be done.
Additionally, have a new interface for the argument passed to the
evdns_callback function. Previously, it was just the original address
we were resolving. But it turns out that, on error, evdns doesn't
tell you the type of the query, so on a failure we didn't know whether
IPv4 or IPv6 queries were failing.
The new convention is to have the first byte of that argument include
the query type. I've refactored the code a bit to make that simpler.
Nick Mathewson [Mon, 5 Nov 2012 18:11:53 +0000 (13:11 -0500)]
Change signature of router_compare_to_my_exit_policy so dns can use it
Also, fix the function so it actually looks at our ipv6 exit policy.
Nick Mathewson [Thu, 1 Nov 2012 02:46:07 +0000 (22:46 -0400)]
Whitespace cleanup
Nick Mathewson [Thu, 1 Nov 2012 02:43:56 +0000 (22:43 -0400)]
Reindent a block in dns.c
Nick Mathewson [Thu, 1 Nov 2012 02:41:20 +0000 (22:41 -0400)]
Actually send BEGIN cell flags
This uses advertised IPv6 ports as an implicit version check.
Nick Mathewson [Thu, 1 Nov 2012 02:40:55 +0000 (22:40 -0400)]
When asking for a specific address type, others aren't acceptable
Nick Mathewson [Thu, 1 Nov 2012 02:18:55 +0000 (22:18 -0400)]
I think it is correct to decorate these addresses.
Nick Mathewson [Thu, 1 Nov 2012 02:04:25 +0000 (22:04 -0400)]
Better checking of exit policies for connections by hostname
Nick Mathewson [Thu, 1 Nov 2012 02:04:09 +0000 (22:04 -0400)]
Only send begin cell flags when we have some to send
Nick Mathewson [Thu, 1 Nov 2012 02:03:54 +0000 (22:03 -0400)]
Never support IPv6 traffic on a SOCKS4 connection.