nihilist/tor
1
0
Fork 0
mirror of https://gitlab.torproject.org/tpo/core/tor.git synced 2024-11-11 21:53:48 +01:00
Code Issues Packages Projects Releases Wiki Activity
17,817 Commits 53 Branches 630 Tags 125 MiB
3b38fd87e8
Commit Graph

456 Commits

Author SHA1 Message Date
Roger Dingledine
be7909f670 Fix a symptom for a bug that straycat found. 
Apparently in rare cases poll() is returning POLLIN for an idle
dns worker. It reads no bytes, presumably because the idle dns
worker didn't write anything in the first place. Gosh, who knows.

Now we only trigger the assert if the idle dns worker actually
tries to send an answer to us.


svn:r2311
 2004-08-25 05:26:09 +00:00
Roger Dingledine
05790d1722 let children survive sigint, sigterm, etc. 
this was biting us because ^c would get delivered to all of them,
maybe because they were all still listening to stdin?


svn:r2197
 2004-08-08 07:25:45 +00:00
Roger Dingledine
3b00c950c2 note a bug in our dns error handling. we should fix this sometime. 
svn:r2149
 2004-08-05 19:33:54 +00:00
Roger Dingledine
aebec8ab9e stop using atexit() to remove our pid, since it's called 
immediately when we daemonize.

also drop our retry period for hidserv desc uploads from 10m to 5m


svn:r2111
 2004-07-22 22:15:36 +00:00
Nick Mathewson
d5e275ddf4 Fix bug reported by peter palfrader: an empty address is unresolvable, but not an occasion for an assertion failure in a dnsworker. 
svn:r2015
 2004-07-06 23:25:21 +00:00
Nick Mathewson
0f3fcfd1db When tor_lookup_hostname says 0.0.0.0, treat it as an error. 
svn:r2003
 2004-07-04 04:52:43 +00:00
Nick Mathewson
3708886939 Implement RESOLVE/RESOLVED cells and socks resolve code 
svn:r1978
 2004-06-17 18:13:09 +00:00
Nick Mathewson
184f4e6044 Rotate dnsworkers and cpuworkers on SIGHUP, so they get new config settings too 
svn:r1950
 2004-06-06 03:38:31 +00:00
Roger Dingledine
69931106f8 be sure to detach streams from the circuit linked list before freeing them 
also, don't bother marking a conn for close if you're about to free it


svn:r1935
 2004-06-02 18:32:24 +00:00
Nick Mathewson
0932505829 Only connection_dns_remove resolving exit connections 
svn:r1934
 2004-06-02 18:12:49 +00:00
Nick Mathewson
0d20fee2fb Add more asserts to dns-pending connections; fix a couple seeming bugs. 
svn:r1931
 2004-06-01 22:09:58 +00:00
Roger Dingledine
f51061439b when the dns resolve is cancelled, or fails, be sure to remove 
conn from circ->resolving_streams

otherwise it gets freed and stays there, causing seg faults.


svn:r1915
 2004-05-21 12:25:15 +00:00
Roger Dingledine
b37450ce55 do all the heavy lifting in connection_about_to_close_connection, 
not in _connection_mark_for_close

this will hopefully clean up the huge cyclical function mess


svn:r1903
 2004-05-19 23:32:20 +00:00
Roger Dingledine
1cf1836990 break connection_edge_end out of connection_mark_for_close 
svn:r1858
 2004-05-12 21:12:33 +00:00
Nick Mathewson
537fb82cbb Make some functions static 
svn:r1857
 2004-05-12 20:58:27 +00:00
Roger Dingledine
3cdf2d67da it's amazing what a bit of punctuation can do for appearances 
svn:r1843
 2004-05-10 10:27:54 +00:00
Roger Dingledine
6168b3222e a few more pieces of markup 
svn:r1841
 2004-05-10 08:56:42 +00:00
Roger Dingledine
341d6f2cab more cleanup, including fleshing out or.h more 
svn:r1839
 2004-05-10 07:37:10 +00:00
Nick Mathewson
81922a125c More doxygenation. 
svn:r1832
 2004-05-10 04:34:48 +00:00
Roger Dingledine
c6d4a00c5f more doxygen markup 
plenty more remains


svn:r1824
 2004-05-09 16:47:25 +00:00
Roger Dingledine
9bf9ca4d09 bugfix: add a circ->resolving_streams field, and make dns resolving streams not actually in the connection_array until they're done resolving. 
bugfix: actually complain if we duplicate mark-for-close a circuit

add more logging for relay ends that claim dns resolve failed, so we can
find out why they're not being retried.


svn:r1798
 2004-05-06 11:08:04 +00:00
Nick Mathewson
682a805092 Comments for nearly all non-tricky files 
svn:r1796
 2004-05-05 21:32:43 +00:00
Nick Mathewson
af08c4f878 Working strerror for windows socket errors, plus some snide comments. 
svn:r1775
 2004-05-02 20:18:21 +00:00
Nick Mathewson
9322641710 Use socketclose on windows as appropriate; end pid files with newline 
svn:r1745
 2004-04-28 21:14:56 +00:00
Roger Dingledine
ca8d50abeb bug resolved (hopefully), so remove comment 
svn:r1738
 2004-04-28 20:12:44 +00:00
Roger Dingledine
22c0cd8a14 clean dns.c better 
svn:r1737
 2004-04-28 20:11:37 +00:00
Roger Dingledine
25d5425fdb fix a bug in dns.c, note but don't fix another one 
svn:r1735
 2004-04-28 19:55:20 +00:00
Nick Mathewson
ac622d94dc Workarounds for a couple of pieces of windows strangeness. 
svn:r1734
 2004-04-28 19:35:12 +00:00
Roger Dingledine
25d54257fb use tor_assert 
remove obsolete BUF_OK macro


svn:r1697
 2004-04-25 20:37:37 +00:00
Nick Mathewson
c024745bd4 Cosmetic patch for arma's fix for The Bug: Don't set the state to 
RESOLVEFAILED if the resolve didn't fail; set it to CONNECTING
instead.


svn:r1590
 2004-04-11 17:07:45 +00:00
Roger Dingledine
d8c954e156 continue beating at pieces of The Bug 
svn:r1588
 2004-04-09 21:31:09 +00:00
Roger Dingledine
23c193c065 if dns resolve is valid, but our exit policy rejects it or 
the connect attempt fails immediately, we were trying to
double-remove it from the dns pending list.

i think this might have been The Bug.
(thanks weasel!)


svn:r1587
 2004-04-09 21:06:14 +00:00
Roger Dingledine
4011906b6b get more serious about hunting The Bug 
(it has been haunting us since the days of 0.0.2preX)


svn:r1576
 2004-04-09 09:39:42 +00:00
Nick Mathewson
3d60cd1d61 Delete trailing whitespace and expand tabs; remove unused aes decrypt code 
svn:r1522
 2004-04-07 19:57:40 +00:00
Roger Dingledine
7118c6177d clean up a few more bugs 
svn:r1365
 2004-03-29 19:57:50 +00:00
Roger Dingledine
0edf0ea7cc make normal exit nodes less likely to crash from the conn-munging bug 
but leave dirservers just as likely to crash


svn:r1363
 2004-03-29 19:51:39 +00:00
Roger Dingledine
4a5586b8e3 make dns resolves a lot faster 
(but only enable this once we've found the bug)


svn:r1358
 2004-03-28 21:16:52 +00:00
Roger Dingledine
f241fdfc73 there is yet another dns-pending-conn-clobbering bug 
somewhere


svn:r1354
 2004-03-28 04:54:36 +00:00
Roger Dingledine
66f878513f deal with hardware word alignment 
this was causing the seg faults on sparc processors

i wonder if i got them all.


svn:r1314
 2004-03-20 09:30:30 +00:00
Roger Dingledine
1ae45b771e don't close all the fd's when you spawn a thread, only when you fork 
svn:r1267
 2004-03-12 21:52:15 +00:00
Roger Dingledine
965a73dc71 bugfix: we were caching transient dns failures 
svn:r1266
 2004-03-12 18:45:42 +00:00
Nick Mathewson
976bacae94 Make all the other read/writes into recv/sends, except when they shouldn't be. 
svn:r1260
 2004-03-11 06:35:03 +00:00
Roger Dingledine
6c9d78b319 stop asserting that computers always go forward in time 
it's simply not true


svn:r1236
 2004-03-06 01:43:37 +00:00
Roger Dingledine
01e06148a1 make workers explain their death better when tor dies 
and make them say it at log level info rather than err


svn:r1232
 2004-03-04 21:26:23 +00:00
Roger Dingledine
8fe504fee6 bugfix: don't pass complex things to the connection_mark_for_close macro 
svn:r1231
 2004-03-04 20:49:38 +00:00
Nick Mathewson
c03e9206ae Clean up dns->connection_mark_for_close->dns_remove path 
svn:r1230
 2004-03-04 18:43:44 +00:00
Roger Dingledine
5701cc2359 fix the function comment 
svn:r1169
 2004-02-28 22:34:47 +00:00
Nick Mathewson
0a673f4491 Split dns_cancel_pending_resolve into dns_cancel_pending_resolve (kill a 
resolve in response to a DNS worker dying) and connection_dns_remove (remove
a pending connection from the resolve structure.)


svn:r1168
 2004-02-28 22:23:44 +00:00
Roger Dingledine
9a55700747 add a log_warn to notify us of yet another potential dns.c bug 
svn:r1167
 2004-02-28 22:13:58 +00:00
Nick Mathewson
833d14246d Make dns_cancel_pending_resolve less agressive. 
svn:r1166
 2004-02-28 22:06:59 +00:00
Nick Mathewson
55174d1cb0 Fix bug on dns_found_answer from mark_to_close stuff. 
svn:r1165
 2004-02-28 21:52:58 +00:00
Roger Dingledine
fadfd26188 remove assert from dns.c that's no longer true: 
typically the resolve will be pending, but it could be failed if it
just failed and we're working through the 'pending' list, or it could
be valid if some other dnsworker raced us to fill in the entry.


svn:r1164
 2004-02-28 21:49:10 +00:00
Nick Mathewson
4e89c6661a Push responsibility for connection marking down as far as possible; have only a close path; add some missing end cells; change return conventions a little. 
svn:r1149
 2004-02-28 04:11:53 +00:00
Nick Mathewson
c5bbb571d2 Refactor mark_for_close, connection_edge_end and friends. Now, everybody 
who wants to shut down a connection calls connection_mark_for_close instead
of setting marked_for_close to 1.  This automatically removes the connection
from the DNS cache if needed, sends a RELAY END cell if appropriate, and can
be changed to do whatever else is needed.

Still to do:
  - The same for circuits, maybe.
  - Add some kind of hold_connection_open_until_flushed flag, maybe.
  - Change stuff that closes connections with return -1 to use mark_for_close,
    maybe.


svn:r1145
 2004-02-27 22:00:26 +00:00
Nick Mathewson
496e414e52 Basic RAM poisoning and magic-checking to notice connection and circuit 
corruption faster; also, check for corruption in dns.c so we can fail fast
for the bug that's nailing Lucky and moria3.


svn:r1123
 2004-02-25 07:31:46 +00:00
Roger Dingledine
8da2e00a83 remove dead code 
svn:r1121
 2004-02-25 06:56:11 +00:00
Roger Dingledine
7985f525ec fix some bugs in assert_connection_ok 
svn:r1118
 2004-02-24 22:33:30 +00:00
Roger Dingledine
598f8b5661 start hunting the lucky-bug 
svn:r1116
 2004-02-22 20:50:20 +00:00
Roger Dingledine
5086300815 split the token bucket into 'rate' and 'burst' params 
we're not entirely migrated to burst yet, for backward compatibility

note some win32 probable-bugs

clean up routerlist.c


svn:r982
 2004-01-10 23:40:38 +00:00
Roger Dingledine
c485725c5a Fix the dns bug: children weren't dying 
We were telling a child to die by closing the parent's file descriptor
to him. But newer children were inheriting the open file descriptor from
the parent, and since they weren't closing them, the socket never closed,
so the child never read eof, so he never knew to exit.

As a side effect to this bug, we were probably failing to properly close
connections to remote hosts, ORs, and OPs, after a dns child was born.

I'm surprised Tor worked at all.


svn:r974
 2004-01-06 07:53:40 +00:00
Roger Dingledine
4a1e05de51 betcha didn't know strncpy could leave an unterminated string 
svn:r949
 2003-12-17 09:20:29 +00:00
Nick Mathewson
4885e90490 Make compile warning-free on cygwin 
svn:r936
 2003-12-15 21:35:52 +00:00
Roger Dingledine
08534a6271 start tracking the 'It appears I've already sent the end' warning 
svn:r932
 2003-12-14 08:32:14 +00:00
Roger Dingledine
7d58bfbd50 change dns notation from question/answer to address/addr 
svn:r929
 2003-12-14 07:40:47 +00:00
Roger Dingledine
05c05d6444 more friendly warning when there are too many dns workers 
svn:r919
 2003-12-14 04:19:12 +00:00
Roger Dingledine
29adf04683 more general cleanup 
svn:r915
 2003-12-14 00:12:02 +00:00
Roger Dingledine
ac56486bf6 use the tor_malloc_zero wrapper 
svn:r837
 2003-11-18 08:20:19 +00:00
Roger Dingledine
56cd147eb9 let getconfig survive repeated calls. now we call it again when we hup. 
change RecommendedVersions into a config option, so dirservers can hup
  for a new one


svn:r809
 2003-11-13 06:49:25 +00:00
Roger Dingledine
aac29076da warn on startup that we don't provide anonymity 
make exit dns cache entries last 15 minutes


svn:r807
 2003-11-13 04:51:34 +00:00
Nick Mathewson
3dddcf4306 Add lots of logging to dns.c; change behavior of often-failing assertion 
svn:r778
 2003-11-08 04:02:05 +00:00
Roger Dingledine
c6b442a346 make end relay cells have payloads 
move default exit policy into config files


svn:r653
 2003-10-22 07:55:44 +00:00
Roger Dingledine
4a66865d0b send the end cell when we realize we're going to end, 
not when we're closing the stream.

this lets us put a payload in the end cell if we want to,
to describe why we're closing the stream.

there are still some places where we don't send the end cell
immediately. i need to track them down. but it's a low priority,
since i've made it send the end cell when we close the stream if
we haven't already sent it.


svn:r640
 2003-10-21 08:37:07 +00:00
Roger Dingledine
9d3f2b232b another minor memory leak 
make dnsconn->address reflect what it's currently resolving


svn:r624
 2003-10-18 07:09:09 +00:00
Nick Mathewson
959b5585a5 Stop using stdout for non-debugging cases 
svn:r592
 2003-10-15 18:28:32 +00:00
Roger Dingledine
b9628f266f change WARNING to WARN 
and fix a few typos


svn:r571
 2003-10-10 01:48:32 +00:00
Roger Dingledine
be874358a4 wrap strdup; prefer time() to gettimeofday() 
svn:r538
 2003-10-04 03:29:09 +00:00
Roger Dingledine
f563bbd2f9 refactor so connection_write_to_buf() never fails 
svn:r537
 2003-10-04 02:38:18 +00:00
Roger Dingledine
cb8212bfcb clean up receiver buckets; prepare for payloads in relay_end; note a few bugs 
svn:r502
 2003-09-27 21:09:56 +00:00
Roger Dingledine
deac704399 connection_new() can't ever fail 
svn:r497
 2003-09-27 07:28:44 +00:00
Roger Dingledine
9e5cafc395 first pass: obey log convention 
ERR is if something fatal just happened

WARNING is something bad happened, but we're still running. The bad thing
is either a bug in the code, an attack or buggy protocol/implementation
of the remote peer, etc. The operator should examine the bad thing and
try to correct it.
(No error or warning messages should be expected. I expect most people
to run on -l warning eventually.)

NOTICE is never ever used.

INFO means something happened (maybe bad, maybe ok), but there's nothing
you need to (or can) do about it.

DEBUG is for everything louder than INFO.


svn:r486
 2003-09-26 10:03:50 +00:00
Nick Mathewson
3d4ccb781a Refactor buffers; implement descriptors. 
'buf_t' is now an opaque type defined in buffers.c .

Router descriptors now include all keys; routers generate keys as
needed on startup (in a newly defined "data directory"), and generate
their own descriptors.  Descriptors are now self-signed.

Implementation is not complete: descriptors are never published; and
upon receiving a descriptor, the directory doesn't do anything with
it.

At least "routers.or" and orkeygen are now obsolete, BTW.


svn:r483
 2003-09-25 05:17:11 +00:00
Roger Dingledine
9c6343fdf8 refactor connects into connection_connect() 
svn:r460
 2003-09-16 01:58:46 +00:00
Roger Dingledine
e585dad887 fix the cpuworker circ-had-vanished bug (maybe) 
still several (many) tls-related bugs outstanding.


svn:r454
 2003-09-14 02:58:50 +00:00
Roger Dingledine
4c7e936a3a note an assert bug that's still around 
svn:r417
 2003-08-25 22:02:42 +00:00
Roger Dingledine
2dda97e8fd implemented cpuworkers 
please poke at it and report bugs

still needs polishing, and only handles onions now (should handle
OR handshakes too)


svn:r402
 2003-08-20 23:05:22 +00:00
Roger Dingledine
88edae9407 start refactoring dnsworker so testing won't be so darn hard 
add NumCpus config variable in preparation for cpuworkers
hardcode /etc/torrc path for config (simplifies win32 port)
improve exit policy debugging during router entry parsing


svn:r397
 2003-08-14 03:52:51 +00:00
Roger Dingledine
e1f2693d05 bugfix: if a dnsworker dies, remember that. 
svn:r396
 2003-08-13 22:38:46 +00:00
Nick Mathewson
2cb4f7a473 Adopt socketpair implementation from perl. For now, always use it. 
svn:r389
 2003-08-12 07:43:15 +00:00
Roger Dingledine
5d9cfb7164 use the spawn_func / spawn_exit abstraction for dnsworkers 
svn:r388
 2003-08-12 07:35:17 +00:00
Nick Mathewson
c336c99e60 Start of port to win32. Missing are: 
- signal support
  - forking for DNS farm
  - changes for async IO
  - daemonizing

In other words, some files still don't build, and the ones that do build,
do nonblocking IO incorrectly.

I'm also not checking in the project files till I have a good place
for them.


svn:r380
 2003-08-12 03:08:41 +00:00
Roger Dingledine
f8203505e1 if stream ends before resolve finishes, inform resolver 
svn:r359
 2003-06-27 00:57:04 +00:00
Roger Dingledine
5d13db862e entries in the dns cache now expire 
(expiry time set to 100 seconds so we can play with it)
exit connections are now informed when pending resolves fail
we kill off the oldest busy worker when we're under attack and need to
  resolve something new


svn:r356
 2003-06-25 07:19:30 +00:00
Roger Dingledine
dbf3435cde simplify fetch_from_buf; cull idle dnsworkers. 
svn:r354
 2003-06-25 00:31:41 +00:00
Roger Dingledine
e061b70dbf a pointer to async-dns alternatives. 
apparently glibc 2.2.4 and windows both have async gethostbynames now.
we'll want to make use of them if they're available (via autoconf).


svn:r347
 2003-06-24 05:17:09 +00:00
Roger Dingledine
05a8c264ff overhaul the dns farm: cut its size in half 
i've eliminated the master dns process, so now the workers just
act like regular connections and are handled by the normal pollarray.

everything seems to still work. ;)


svn:r327
 2003-06-17 14:31:05 +00:00
Roger Dingledine
d3292e484a implement truncate and truncated (untested) 
clean up circuit_deliver_relay_cell convention


svn:r312
 2003-06-12 10:16:33 +00:00
Roger Dingledine
39e9d79038 add circuit-level sendme relay cells 
remove sendme cells
replace malloc with tor_malloc
patch (but not track down) bug in onion pending list
streamline connection_ap handshake


svn:r293
 2003-05-20 06:41:23 +00:00
Nick Mathewson
c217b1ca71 Add magic to end of C files to make emacs happy; split test invocation into separate file. 
svn:r224
 2003-04-07 02:12:02 +00:00
Roger Dingledine
afa95ca47c make servers tolerate the dns bug. 
svn:r223
 2003-04-05 19:53:45 +00:00
Roger Dingledine
1768f29d7e better comments and a few patches 
svn:r164
 2003-03-04 04:36:37 +00:00
Roger Dingledine
1714ea2ab8 Bugfixes and enhancements in sendmes and dns farm 
svn:r161
 2003-02-18 01:35:55 +00:00
Roger Dingledine
8c136b3a06 bugfix: actually remember the answer in the cache :) 
svn:r159
 2003-02-16 02:05:24 +00:00
Roger Dingledine
3cf02a88f4 rudimentary dns caching (of both resolves and resolve failures) 
serious performance increase over non-caching


svn:r158
 2003-02-14 07:53:55 +00:00
Roger Dingledine
0bc8dc1314 fix endian issues for topics -- they might work on bsd now 
(they wouldn't have before)

alternate code which bypasses the dns farm, so we can compare speed


svn:r154
 2003-02-06 23:48:35 +00:00
Roger Dingledine
ceafe12ed6 make reusing circuits work (and be the default) 
performance is better, but not by much. not sure why yet.


svn:r153
 2003-02-06 08:00:49 +00:00
Roger Dingledine
c35373a2cf major overhaul: dns slave subsystem, topics 
on startup, it forks off a master dns handler, which forks off dns
slaves (like the apache model). slaves as spawned as load increases,
and then reused. excess slaves are not ever killed, currently.

implemented topics. each topic has a receive window in each direction
at each edge of the circuit, and sends sendme's at the data level, as
per before. each circuit also has receive windows in each direction at
each hop; an edge sends a circuit-level sendme as soon as enough data
cells have arrived (regardless of whether the data cells were flushed
to the exit conns). removed the 'connected' cell type, since it's now
a topic command within data cells.

at the edge of the circuit, there can be multiple connections associated
with a single circuit. you find them via the linked list conn->next_topic.

currently each new ap connection starts its own circuit, so we ought
to see comparable performance to what we had before. but that's only
because i haven't written the code to reattach to old circuits. please
try to break it as-is, and then i'll make it reuse the same circuit and
we'll try to break that.


svn:r152
 2003-01-26 09:02:24 +00:00