Commit Graph

458 Commits

Author SHA1 Message Date
Nick Mathewson
df9c8feac7 Commit fixes for several pending tor core tasks: document all DOCDOCed functions; time out uncontrolled unattached streams; feed reasons to SOCKS5 (refactoring connection_ap_handshake_socks_reply in the process); change DirFetchPeriod/StatusFetchPeriod to have a special "Be smart" value.
svn:r3769
2005-03-17 12:38:37 +00:00
Nick Mathewson
0197b47ce9 Fix some more obscure compiler warnings
svn:r3758
2005-03-14 03:28:46 +00:00
Nick Mathewson
97dad670ea Renormalize whitespace
svn:r3757
2005-03-14 03:18:35 +00:00
Nick Mathewson
777d3dde0c Get address map resetting implemented.
svn:r3745
2005-03-11 21:39:39 +00:00
Nick Mathewson
e14f812a67 Still more code to make sure we send the right number and kind of RELAY END cells
svn:r3723
2005-03-02 03:13:05 +00:00
Nick Mathewson
349ee1abea Clean up rend cache on shutdown
svn:r3714
2005-02-28 22:38:00 +00:00
Nick Mathewson
97bc49bd72 Try a little harder to avoid openssl SSL* double-free reports.
svn:r3710
2005-02-28 02:52:51 +00:00
Roger Dingledine
b3177d509e free some more memory at shutdown
svn:r3707
2005-02-28 01:59:18 +00:00
Roger Dingledine
a1f1fa6ab2 Checkpoint in-progress fixes:
Add 'testing' circuit purpose, for reachability testing.
Notice when our IP changes, and reset stats.
Try to pull down a directory via Tor to see if our DirPort is working.
Try to extend a circuit back to us to see if our ORPort is working.
Only publish a descriptor if they're both reachable.

These mostly work, and I'd better get them in before I cause conflicts.


svn:r3703
2005-02-27 09:47:01 +00:00
Nick Mathewson
26e6eb7539 Check returns from libevent calls better; stop poking at internals of events to see if they are live.
svn:r3691
2005-02-25 05:42:01 +00:00
Nick Mathewson
2ff6cfccd6 Forward-port fix to make dual-purpose NT executable work.
svn:r3665
2005-02-23 05:34:25 +00:00
Roger Dingledine
55c3f15ed1 Forward-port the trick to bump up ConnLimit as far as it can
go, but not past MAXCONNECTIONS. Also throw away the FAKE_POLL
caveats since 0.1.0 doesn't need it any more.


svn:r3657
2005-02-22 23:52:44 +00:00
Nick Mathewson
1a55422006 Fix windows build: juggle }s and #endifs around in main.c
svn:r3645
2005-02-22 02:57:43 +00:00
Roger Dingledine
8727acf2f1 Add a new AddressMap directive to rewrite incoming socks addresses.
Add a new TrackHostExits directive to trigger addressmaps for
certain incoming socks addresses, for sites that break when your exit
keeps changing.
Redo the client-side dns cache so it's just an addressmap too.


svn:r3641
2005-02-22 00:53:08 +00:00
Nick Mathewson
5d836c8140 Free tls resources on exit too
svn:r3615
2005-02-11 01:41:19 +00:00
Nick Mathewson
f672577bc6 Free even more things on shutdown. Temporarily move tor_free_all out from #ifdef so it gets tested more.
svn:r3614
2005-02-11 01:26:47 +00:00
Nick Mathewson
60f234f42b Add more functions to free things to help dmalloc allong.
svn:r3613
2005-02-10 23:18:39 +00:00
Roger Dingledine
c5c46d6fb6 don't list version in the win32 service display name or service name,
since they'll keep changing and people will get messed up about which
one is installed.


svn:r3592
2005-02-09 02:26:24 +00:00
Nick Mathewson
bb32fa86d7 Re-normalize whitespace; make "make check-spaces" happy
svn:r3566
2005-02-05 21:42:46 +00:00
Nick Mathewson
24ad6542c9 forward-port NT service patch
svn:r3565
2005-02-05 21:40:25 +00:00
Roger Dingledine
bcece3d8a7 another minor fix
svn:r3511
2005-02-03 07:30:00 +00:00
Peter Palfrader
ddd724ef94 Add --with-dmalloc configure option
svn:r3508
2005-02-03 07:25:10 +00:00
Peter Palfrader
dfa3a0220f Write a function header for Roger so he can fill it with life
svn:r3507
2005-02-03 07:23:48 +00:00
Roger Dingledine
0f5a198028 sometimes, a conn on the closeable_lst gets absolutely
clobbered. maybe this will help me find out why.


svn:r3498
2005-02-02 00:19:09 +00:00
Roger Dingledine
2e97593d25 define TOR_FRAGILE if you want tor to give you a core when
something goes wrong. this should only be used by people actively
tracking bugs.


svn:r3487
2005-02-01 00:37:16 +00:00
Roger Dingledine
d6489d8e41 lay a trap for situations where we connection_free something
that's still on one of the lists.

in particular, this will detect situations where we mark-for-close
and then free it ourselves.


svn:r3475
2005-01-31 00:33:32 +00:00
Roger Dingledine
75b45bcd4d stop the infinite loop of freeing the same connection_t over and
over when it's not linked into the connection_array


svn:r3474
2005-01-31 00:26:09 +00:00
Roger Dingledine
3969b6066e Stop crashing when we're asking to close_if_marked a conn that
hasn't been connection_added yet. This happens when an exit conn
is in dns_wait and we get a relay end cell for it before it finishes.

We were silently leaking each of these marked conns in 0.0.9.x. Now
we actually free them.


svn:r3470
2005-01-30 21:47:47 +00:00
Roger Dingledine
7ddeb7fae7 try to find the corrupt closeable_connection_lst entry early
svn:r3468
2005-01-30 20:31:08 +00:00
Roger Dingledine
8e7fc8d259 comment about a bug wrt clock skew
svn:r3461
2005-01-30 00:07:31 +00:00
Roger Dingledine
b2fbd834f0 forward-port the dns and maxconn fixes
svn:r3448
2005-01-28 08:53:47 +00:00
Nick Mathewson
e03f571ee3 Forward-port win32 service patch; try to clean it a touch.
svn:r3438
2005-01-27 21:45:32 +00:00
Nick Mathewson
6704093c0d Appease another 64-bit warning
svn:r3392
2005-01-20 20:21:34 +00:00
Nick Mathewson
5b45527851 Appease weasel's paranoid 64-bit compiler.
svn:r3389
2005-01-20 20:16:48 +00:00
Roger Dingledine
d2400a5afd Introduce a notion of 'internal' circs, which are chosen without regard
to the exit policy of the last hop. Intro and rendezvous circs must
be internal circs, to avoid leaking information. Resolve and connect
streams can use internal circs if they want.

New circuit pooling algorithm: make sure to have enough circs around
to satisfy any predicted ports, and also make sure to have 2 internal
circs around if we've required internal circs lately (with high uptime
if we've seen that lately).

Split NewCircuitPeriod config option into NewCircuitPeriod (30 secs),
which describes how often we retry making new circuits if current ones
are dirty, and MaxCircuitDirtiness (10 mins), which describes how long
we're willing to make use of an already-dirty circuit.

Once rendezvous circuits are established, keep using the same circuit as
long as you attach a new stream to it at least every 10 minutes. (So web
browsing doesn't require you to build new rend circs every 30 seconds.)

Cannibalize GENERAL circs to be C_REND, C_INTRO, S_INTRO, and S_REND
circ as necessary, if there are any completed ones lying around when
we try to launch one.

Re-instate the ifdef's to use version-0 style introduce cells, since
there was yet another bug in handling version-1 style. We'll try switching
over again after 0.0.9 is obsolete.

Bugfix: when choosing an exit node for a new non-internal circ, don't take
into account whether it'll be useful for any pending x.onion addresses --
it won't.

Bugfix: we weren't actually publishing the hidden service descriptor when
it became dirty. So we only published it every 20 minutes or so, which
means when you first start your Tor, the hidden service will seem broken.


svn:r3360
2005-01-17 18:13:09 +00:00
Nick Mathewson
a38bccdca7 Move call to event_init() immediately after start_daemon(); move start_daemon earlier. This should fix using libevent with RunAsDaemon.
svn:r3354
2005-01-13 21:32:08 +00:00
Roger Dingledine
abea142665 more cleanup
svn:r3347
2005-01-12 12:22:25 +00:00
Roger Dingledine
24a97d2c55 fix up some comments
svn:r3345
2005-01-12 12:18:46 +00:00
Roger Dingledine
3809c0ab91 tolerate bandwidtch buckets going negative (i hope)
svn:r3344
2005-01-12 10:00:38 +00:00
Nick Mathewson
324b192f68 Make Tor use Niels Provos's libevent instead of it's current
poll-but-sometimes-select mess.  This will let us use faster async cores
(like epoll, kpoll, and /dev/poll), and hopefully work better on Windows
too.

There are some fairly nasty changes to main.c here; this will almost
certainly break something.  But hey, that's what alphas are for.


svn:r3341
2005-01-12 06:42:32 +00:00
Roger Dingledine
70b3208f0a if somebody starts his tor server in jan 2004 and then fixes his
clock, don't make his published uptime be a year.


svn:r3337
2005-01-11 11:54:11 +00:00
Roger Dingledine
4b6d55afdf separate time_of_process_start from stats_n_seconds_uptime:
now we reset uptime when a dir fetch entirely fails. this
hopefully has something to do with stability of being on the
network.


svn:r3333
2005-01-10 05:10:22 +00:00
Roger Dingledine
3370f53010 it helps to use the correct name for the struct
svn:r3331
2005-01-10 04:32:59 +00:00
Roger Dingledine
10a81ccf18 when pruning circuit-wait streams when a dir fetch failed,
refer in logs to the destination address, not the address of the
application connection.


svn:r3330
2005-01-10 02:47:50 +00:00
Roger Dingledine
f468ff04ea when we haven't fetched a directory yet, or the last time we
tried they were all unreachable, assume we are not connected to
the network.

when an application request comes in during this state, be
optimistic and assume we just reconnected. fetch a new directory
and if it works, begin making circuits.


svn:r3327
2005-01-07 15:57:57 +00:00
Roger Dingledine
f8b517fa97 finish marking my comment in main.c
svn:r3318
2005-01-06 19:20:54 +00:00
Roger Dingledine
1f51809e65 clean up string printing in relay logs
svn:r3317
2005-01-06 19:19:13 +00:00
Nick Mathewson
dcd228585d Implement SIGNAL control command.
svn:r3307
2005-01-05 06:40:47 +00:00
Nick Mathewson
0c4186d6ee Use correct logic to decide whether there are more directory servers to retry.
svn:r3257
2005-01-03 20:51:24 +00:00
Roger Dingledine
e1ac30c2bb stop calling assert_connection_ok(conn, time(NULL)) on every single
conn on every single loop.


svn:r3229
2004-12-25 06:43:40 +00:00