Commit Graph

11335 Commits

Author SHA1 Message Date
Roger Dingledine
98ce45bebf Backport candidate:
Fix a potential crash bug when we load many server descriptors at
once and some of them make others of them obsolete. Fixes bug
458. [Bugfix on 0.1.2.x]


svn:r10832
2007-07-16 00:11:03 +00:00
Roger Dingledine
5ab3f49d93 backport candidate: fix a memory leak from r9039 which had us leaking
50ish bytes every time we rebuild the cached-router files.


svn:r10830
2007-07-15 08:34:32 +00:00
Roger Dingledine
2c2f7ee6a4 fix a memory leak from r10504 (i think)
svn:r10829
2007-07-15 08:11:56 +00:00
Andrew Lewman
76c5fbfe28 eventdns seems to run fine on osx, rpm-based linuxes, and win32.
Include it in the build instructions.


svn:r10826
2007-07-14 02:32:19 +00:00
Nick Mathewson
db8670f58a svn:r10825 2007-07-13 16:14:08 +00:00
Roger Dingledine
8ad4b8a640 some cleanups on phrasing in proposal 114
svn:r10821
2007-07-12 18:06:13 +00:00
Peter Palfrader
2ce1cb22bd Comment one tor_assert
svn:r10820
2007-07-12 17:37:47 +00:00
Roger Dingledine
ef6bc1a451 Keep streamids from different exits on a circuit separate. This
bug may have allowed other routers on a given circuit to inject
cells into streams. Reported by lodger; fixes bug 446. [Bugfix
on 0.1.2.x]


svn:r10818
2007-07-12 17:09:19 +00:00
Nick Mathewson
656b7761a8 r13730@catbus: nickm | 2007-07-12 12:32:40 -0400
Patch from lodger: avoid roundoff-error-induced crash bugs when picking routers by bandwidth.
 Also, remove listed backports for 0.1.2.x; that list is now in TODO.012


svn:r10812
2007-07-12 16:34:45 +00:00
Peter Palfrader
9ed261cd6a In router_add_to_routerlist() only call router_have_minimum_dir_info() under some circumstances, not always
svn:r10799
2007-07-12 15:57:23 +00:00
Roger Dingledine
6ade658732 admit that our rend spec was lying. sorry karsten!
svn:r10798
2007-07-12 15:27:52 +00:00
Roger Dingledine
706fad08fa vague fuzzy bug reports so i don't forget them
svn:r10797
2007-07-12 15:24:56 +00:00
Roger Dingledine
9bf23e0598 backport candidate:
When sending destroy cells from a circuit's origin, don't include
the reason for tearing down the circuit. The spec says we didn't,
and now we actually don't. Reported by lodger. [Bugfix on 0.1.2.x]


svn:r10790
2007-07-11 03:37:45 +00:00
Roger Dingledine
83e8d8d9b9 clean up a comment
svn:r10789
2007-07-11 03:28:12 +00:00
Nick Mathewson
39dc589d32 r13691@catbus: nickm | 2007-07-10 18:44:30 -0400
Fix an assert error in 455 patch.


svn:r10788
2007-07-10 22:44:32 +00:00
Nick Mathewson
fa1376014b r13689@catbus: nickm | 2007-07-10 17:46:30 -0400
tweaks to last patch.  from croup.


svn:r10787
2007-07-10 21:46:33 +00:00
Nick Mathewson
7e20fdbefb r13687@catbus: nickm | 2007-07-10 16:08:14 -0400
Possible partial fix for bug 455: use eos logic everywhere.


svn:r10786
2007-07-10 20:08:18 +00:00
Nick Mathewson
d6ba1c3d96 r13676@catbus: nickm | 2007-07-10 13:41:24 -0400
Man. I thought I removed that code.


svn:r10785
2007-07-10 17:41:26 +00:00
Nick Mathewson
4325fc5e83 r13674@catbus: nickm | 2007-07-10 13:27:30 -0400
Re-wrap proposal 117 so it fits in 80 columns.


svn:r10784
2007-07-10 17:27:33 +00:00
Nick Mathewson
81083cf0ce r13672@catbus: nickm | 2007-07-10 13:17:08 -0400
Add IPv6 exit proposal from coderman.


svn:r10783
2007-07-10 17:17:14 +00:00
Nick Mathewson
a8465adaa2 r13667@catbus: nickm | 2007-07-10 13:12:52 -0400
Patch from tup: add GETINFO status/enough-dir-info


svn:r10782
2007-07-10 17:14:55 +00:00
Nick Mathewson
74711479eb r13666@catbus: nickm | 2007-07-10 13:10:00 -0400
Get the RESOLVE controller code working.


svn:r10781
2007-07-10 17:14:51 +00:00
Nick Mathewson
7705314777 r13637@catbus: nickm | 2007-07-06 14:24:27 -0400
fix copy-paste error in config message src/or/config.c .  Apply
 initial (but pared-down) version of launch-resolve-via-controller
 patch from Robert Hogan.
 


svn:r10780
2007-07-10 17:13:24 +00:00
Karsten Loesing
ef246793d3 incorporated suggestions by Roger, added status of specification and implementation for upcoming GSoC mid-term evaluation
svn:r10773
2007-07-09 14:54:48 +00:00
Nick Mathewson
10e5ed1a84 r13602@Kushana: nickm | 2007-07-07 23:44:24 -0400
Backport candidate: fix a possible buffer overrun when using natd support on old BSDs.  Based on a patch from "Mr. Croup".


svn:r10760
2007-07-08 03:45:47 +00:00
Peter Palfrader
db3a4bf897 Call router_have_minimum_dir_info() at the beginning of router_add_to_routerlist() since it has side effects
svn:r10754
2007-07-06 21:21:09 +00:00
Peter Palfrader
39b1a11f8d Remove two arguments from routerlist_replace(). idx, which is
the old router's index in the routerlist, can be gotten from the
old routerinfo itself anyway, and make_old always was true.
(There only is one caller of routerlist_replace())


svn:r10753
2007-07-06 18:40:39 +00:00
Nick Mathewson
e6ba07f8f5 r13634@catbus: nickm | 2007-07-06 14:20:50 -0400
Fix a crash bug (*the* crash bug?) on authorities.  Backport candidate.


svn:r10752
2007-07-06 18:20:54 +00:00
Nick Mathewson
92b6e09082 r13633@catbus: nickm | 2007-07-06 11:08:19 -0400
Call tor_init_threads() from tor_init().


svn:r10751
2007-07-06 18:20:52 +00:00
Nick Mathewson
73b4428a8b r13631@catbus: nickm | 2007-07-06 10:17:22 -0400
Try to fix win32 build again.


svn:r10750
2007-07-06 14:17:30 +00:00
Nick Mathewson
bbc7cf86d5 r13600@Kushana: nickm | 2007-07-05 23:49:47 -0400
try to fix mingw compile


svn:r10743
2007-07-06 03:49:57 +00:00
Peter Palfrader
a1e3b60d2d Pass arguments given to autogen on to configure. Let's hope this doesn't break things much
svn:r10742
2007-07-05 21:52:11 +00:00
Nick Mathewson
a9469098ca r13596@kushana: nickm | 2007-07-05 10:53:22 -0400
Argh, re-enable CRITICAL_SECTION code in win32.


svn:r10741
2007-07-05 14:53:27 +00:00
Nick Mathewson
5c9c420111 svn:r10740 2007-07-05 14:51:10 +00:00
Nick Mathewson
ae4ab0f617 r13454@kushana: nickm | 2007-06-20 14:22:44 -0400
Switch windows locking implementation to CRITICAL_SECTION, not Mutex: Mutex is heavier-weight, and meant for multi-process situations.


svn:r10739
2007-07-05 14:51:01 +00:00
Nick Mathewson
04c9ae8164 r13453@kushana: nickm | 2007-06-20 10:01:37 -0400
Fix a type error in the last checkin.


svn:r10738
2007-07-05 14:50:13 +00:00
Nick Mathewson
b5c8a6183a r13452@kushana: nickm | 2007-06-20 09:52:55 -0400
Patch from Robert Hogan: Generate STREAM NEW events for dnsport requests and tunneled directory connections.  Log when we are testing for hijacking.


svn:r10737
2007-07-05 14:50:05 +00:00
Nick Mathewson
92b0fc2e53 r13451@kushana: nickm | 2007-06-20 09:43:26 -0400
Fix typos and clarify points in dir-spec. Patch from Karsten.


svn:r10736
2007-07-05 14:49:57 +00:00
Nick Mathewson
3e75cf7b57 svn:r10735 2007-07-05 14:49:27 +00:00
Roger Dingledine
99bfb1bebe If there's a never-before-connected-to guard node in our list,
never choose any guards past it. This way we don't expand our 
guard list unless we need to. [Bugfix in 0.1.2.x]

I'm not sure if this will solve all our problems, but it is at least
something.



svn:r10730
2007-07-02 22:15:26 +00:00
Roger Dingledine
5c1daed3ad the non-controversial changes i found in my sandbox
svn:r10729
2007-07-02 22:07:53 +00:00
Nick Mathewson
32b165ab61 r13604@catbus: nickm | 2007-07-02 16:58:40 -0400
Some temporary output to try to figure out where bsd thread unit tests are going wrong


svn:r10728
2007-07-02 21:08:03 +00:00
Peter Palfrader
64f4cff192 I so wonder how this blows up on the real network - make _routerlist_find_elt be strict about the idx it is passed - if it is not -1 then it has to be correct
svn:r10727
2007-07-02 20:17:12 +00:00
Peter Palfrader
d071df748a If we want to be more strict in _routerlist_find_elt I better call it properly
svn:r10726
2007-07-02 20:04:26 +00:00
Nick Mathewson
b092f7f8d2 r13594@catbus: nickm | 2007-07-02 15:53:35 -0400
dir-spec.txt clarifications suggested by Lee Fisher.


svn:r10725
2007-07-02 19:53:57 +00:00
Peter Palfrader
ccefd7404e Don't get private with smartlists when there's a perfectly fine accessor method for num_used
svn:r10724
2007-07-02 19:45:44 +00:00
Peter Palfrader
f274c1413e Make it build with -O0 as well
svn:r10722
2007-07-02 17:50:57 +00:00
Peter Palfrader
adff891463 Make DEBUG_SMARTLIST work
svn:r10721
2007-07-02 17:48:56 +00:00
Peter Palfrader
ea1348360e Three more assertions which hopefully are sane, Nick? They might help track down #417.
svn:r10720
2007-07-02 15:13:52 +00:00
Nick Mathewson
1bdcfd9203 r13570@catbus: nickm | 2007-06-30 20:41:05 -0400
Implement conditions in compat.c; switch windows to use "critical sections" instead of mutexes.  Apparently, mutexes are for IPC and critical sections are for multithreaded.


svn:r10716
2007-07-01 16:22:45 +00:00