Nick Mathewson
60880cda84
Resolve a bunch of FIXME items; mark a lot more for attention; ask for clarification on some. Turn all XXXX008 ("showstopper for 0.0.8 release") items into XXXX009 or XXXX, since plainly they were not showstoppers for 0.0.8. Add/clean some docs.
...
svn:r2808
2004-11-12 16:39:03 +00:00
Roger Dingledine
394554cfbf
Clean up copyrights.
...
Break connection_consider_empty_buckets() out of
connection_read_bucket_decrement().
svn:r2698
2004-11-07 01:33:06 +00:00
Roger Dingledine
6c61ed4fb5
Make options no longer a global variable.
...
Now we can try setting an option but back out if it fails to parse, or
if it's disallowed (e.g. changing RunAsDaemon from 1 to 0).
Use parse_line_from_str rather than parse_line_from_file.
svn:r2692
2004-11-06 05:18:11 +00:00
Nick Mathewson
6874c39904
Fix windows build
...
svn:r2613
2004-10-27 18:14:38 +00:00
Nick Mathewson
6980929e64
Use strlcpy, not strcpy.
...
svn:r2610
2004-10-27 06:48:16 +00:00
Nick Mathewson
f67f83b1fa
Use strlcpy, not strncpy
...
svn:r2603
2004-10-27 06:03:28 +00:00
Roger Dingledine
de65052312
don't assert multiple things in the same tor_assert()
...
svn:r2544
2004-10-16 22:14:52 +00:00
Nick Mathewson
a42adce362
fix memory leak in router.c; start relying on NULL==(zero bytes)
...
svn:r2538
2004-10-16 20:38:57 +00:00
Roger Dingledine
aebc3a03ba
more int to size_t conversions, fixing one or more amd64 bugs
...
plus a whitespace patch on config.c from vicman
svn:r2482
2004-10-14 02:47:09 +00:00
Roger Dingledine
013104a1e2
people are starting to bump into the limit of 50 dns workers.
...
raise it to 100 for now. we need a better solution long-term.
but alas, there are no better solutions that are also portable.
hm.
svn:r2363
2004-09-22 03:56:41 +00:00
Nick Mathewson
9defe8a307
Attach dummy resolve connections to a circuit *before* calling dns_resolve(). This fixes a bug where cached answers would never be sent in RESOLVED cells.
...
svn:r2360
2004-09-21 18:12:12 +00:00
Roger Dingledine
d977677aa8
bugfix: when a resolve fails for a begin request, actually send
...
the end cell back. also, give a better reason for failure.
svn:r2358
2004-09-21 17:33:05 +00:00
Nick Mathewson
0dcc8c7048
Bugfix: When a dns request failed, we would detach any "resolving" streams from the circuit before we would send a the resolve_failed cells.
...
svn:r2355
2004-09-21 16:42:07 +00:00
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