Commit Graph

1123 Commits

Author SHA1 Message Date
Sebastian Hahn
e0c3de40ad Fix check-spaces complaints 2015-02-06 21:36:40 +01:00
Nick Mathewson
d1e52d9a2a Correctly handle OutboundBindAddress again.
ca5ba2956b broke this; bug not in any
released Tor.

Also fix a typo.

Fixes 14541 and 14527.  Reported by qbi.
2015-01-30 07:29:23 -05:00
Nick Mathewson
4c1a779539 Restrict unix: addresses to control and socks for now 2015-01-29 14:51:59 -05:00
Nick Mathewson
204374f7d9 Remove SocksSocket; it's now spelled differently thanks to 14451
Also, revise bug12585 changes file to mention new syntax
2015-01-29 14:46:20 -05:00
Andrea Shepard
ca5ba2956b Support connection_exit_connect() to AF_UNIX sockets 2015-01-28 14:30:23 -05:00
Nick Mathewson
23fc1691b6 Merge branch 'better_workqueue_v3_squashed' 2015-01-21 14:47:16 -05:00
Nick Mathewson
4b23b398a3 Merge branch 'bug8546_squashed'
Conflicts:
	src/or/connection.c
	src/or/or.h
	src/or/relay.c
2015-01-16 09:31:50 -05:00
Nick Mathewson
49bdfbabb4 Replace field-by-field copy with memcpy for entry_port_cfg 2015-01-16 09:23:03 -05:00
Nick Mathewson
13dac5e463 Move entry_port_cfg_t fields in entry_connection_t
Also rename some options for uniformity, and apply this script:

@@
entry_connection_t *conn;
@@
 conn->
+entry_cfg.
\(
 isolation_flags
\|
 session_group
\|
 socks_prefer_no_auth
\|
 ipv4_traffic
\|
 ipv6_traffic
\|
 prefer_ipv6
\|
 cache_ipv4_answers
\|
 cache_ipv6_answers
\|
 use_cached_ipv4_answers
\|
 use_cached_ipv6_answers
\|
 prefer_ipv6_virtaddr
\)
2015-01-16 09:22:58 -05:00
Nick Mathewson
58d17add5e Combine entry_port_cfg_t fields in listener_connection_t
Also, revise the code using these options with this cocci script:

@@
listener_connection_t *conn;
@@
 conn->
+entry_cfg.
\(
 isolation_flags
\|
 session_group
\|
 socks_prefer_no_auth
\|
 ipv4_traffic
\|
 ipv6_traffic
\|
 prefer_ipv6
\|
 cache_ipv4_answers
\|
 cache_ipv6_answers
\|
 use_cached_ipv4_answers
\|
 use_cached_ipv6_answers
\|
 prefer_ipv6_virtaddr
\)
2015-01-16 09:22:53 -05:00
Nick Mathewson
f444f2b1d3 Split client-specific and server-specific parts of port_cfg_t
Also, apply this cocci script to transform accesses. (Plus manual
migration for accesses inside smartlist_foreach loops.)

@@

port_cfg_t *cfgx;
@@
 cfgx->
+server_cfg.
\(
 no_advertise
\|
 no_listen
\|
 all_addrs
\|
 bind_ipv4_only
\|
 bind_ipv6_only
\)

@@

port_cfg_t *cfgx;
@@
 cfgx->
+entry_cfg.
\(
 isolation_flags
\|
 session_group
\|
 socks_prefer_no_auth
\|
 ipv4_traffic
\|
 ipv6_traffic
\|
 prefer_ipv6
\|
 cache_ipv4_answers
\|
 cache_ipv6_answers
\|
 use_cached_ipv4_answers
\|
 use_cached_ipv6_answers
\|
 prefer_ipv6_virtaddr
\)
2015-01-16 09:22:49 -05:00
Nick Mathewson
1e896214e7 Refactor cpuworker to use workqueue/threadpool code. 2015-01-14 11:23:34 -05:00
Nick Mathewson
518b0b3c5f Do not log a notice on every socks connection 2015-01-14 09:54:40 -05:00
Nick Mathewson
17c568b95c Fix new unused variable warning in connection_listener_new 2015-01-13 13:45:35 -05:00
Andrea Shepard
066acaf6b9 Explicitly chmod AF_UNIX sockets to 0600 when *GroupWritable isn't specified 2015-01-13 00:27:04 +00:00
Andrea Shepard
4316bb601a Remove no-longer-accurate comment from connection.c 2015-01-13 00:21:59 +00:00
Andrea Shepard
62f297fff0 Kill duplicated code in connection_listener_new() 2015-01-12 16:26:34 +00:00
Andrea Shepard
a3bcde3638 Downgrade open/close log message for SocksSocket 2015-01-07 22:57:51 +00:00
Andrea Shepard
2ca1c386b0 Bring sanity to connection_listener_new() 2015-01-07 22:51:24 +00:00
Andrea Shepard
48633c0766 Rename is_tcp in connection_listener_new(), since AF_UNIX means SOCK_STREAM no longer implies TCP 2015-01-07 19:45:59 +00:00
Andrea Shepard
c6451e4c9f Refactor check_location_for_unix_socket()/check_location_for_socks_unix_socket() to eliminate duplicated code 2015-01-07 19:17:04 +00:00
Jacob Appelbaum
8d59ddf3cb Commit second draft of Jake's SOCKS5-over-AF_UNIX patch. See ticket #12585.
Signed-off-by: Andrea Shepard <andrea@torproject.org>
2015-01-07 17:42:57 +00:00
Nick Mathewson
1abd526c75 Merge remote-tracking branch 'public/bug12985_025' 2015-01-07 11:55:50 -05:00
Nick Mathewson
f54e54b0b4 Bump copyright dates to 2015, in case someday this matters. 2015-01-02 14:27:39 -05:00
Nick Mathewson
1c05dfd0b6 Merge branch 'ticket7356_squashed' 2014-12-21 14:48:53 -05:00
rl1987
fc7d5e598b Using CHANNEL_FINISHED macro in connection.c 2014-12-21 14:48:38 -05:00
Nick Mathewson
a28df3fb67 Merge remote-tracking branch 'andrea/cmux_refactor_configurable_threshold'
Conflicts:
	src/or/or.h
	src/test/Makefile.nmake
2014-11-27 22:39:46 -05:00
Nick Mathewson
fcdcb377a4 Add another year to our copyright dates.
Because in 95 years, we or our successors will surely care about
enforcing the BSD license terms on this code.  Right?
2014-10-28 15:30:16 -04:00
rl1987
14d59fdc10 Updating message that warns about running out of sockets we can use. 2014-10-28 14:13:25 -04:00
Andrea Shepard
8852a1794c Track total queue size per channel, with overhead estimates, and global queue total 2014-09-30 22:49:03 -07:00
Nick Mathewson
b448ec195d Clear the cached address from resolve_my_address() when our IP changes
Closes 11582; patch from "ra".
2014-09-29 13:47:58 -04:00
Nick Mathewson
e07206afea Merge remote-tracking branch 'yawning/bug_8402' 2014-09-10 23:41:55 -04:00
Nick Mathewson
e3c143f521 Merge remote-tracking branch 'origin/maint-0.2.5' 2014-09-02 11:58:08 -04:00
Nick Mathewson
efcab43956 Fix a number of clang analyzer false-positives
Most of these are in somewhat non-obvious code where it is probably
a good idea to initialize variables and add extra assertions anyway.

Closes 13036.  Patches from "teor".
2014-09-02 11:56:56 -04:00
Nick Mathewson
d8fe499e08 Revert "restore the sensible part of ac268a83408e1450544db2f23f364dfa3"
This reverts commit b82e166bec.

We don't need that part in 0.2.5, since 0.2.5 no longer supports
non-multithreaded builds.
2014-08-29 12:25:05 -04:00
Nick Mathewson
b0138cd055 Merge remote-tracking branch 'public/bug12985_024' into bug12984_025 2014-08-29 12:24:52 -04:00
Nick Mathewson
b82e166bec restore the sensible part of ac268a8340
We don't want to call event_del() postfork, if cpuworkers are
multiprocess.
2014-08-29 12:21:57 -04:00
Nick Mathewson
4144b4552b Always event_del() connection events before freeing them
Previously, we had done this only in the connection_free() case, but
when we called connection_free_() directly from
connections_free_all(), we didn't free the connections.
2014-08-29 11:33:05 -04:00
Nick Mathewson
b408125288 Merge remote-tracking branch 'andrea/bug11302' 2014-07-16 16:58:41 +02:00
Anthony G. Basile
d504a4e36f src/or/connection.c: expose bucket_millis_empty for bufferevents test
Currently tor fails to build its test when enabled with bufferevents
because an #ifndef USE_BUFFEREVENTS hides bucket_millis_empty() and
friends.  This is fine if we don't run tests, but if we do, we need
these functions in src/or/libtor-testing.a when linking src/test/test.

This patch moves the functions outside the #ifndef and exposes them.

See downstream bug:

	https://bugs.gentoo.org/show_bug.cgi?id=510124
2014-07-16 10:37:00 +02:00
Nick Mathewson
3a2e25969f Merge remote-tracking branch 'public/ticket6799_024_v2_squashed'
Conflicts:
	src/or/channel.c
	src/or/circuitlist.c
	src/or/connection.c

Conflicts involved removal of next_circ_id and addition of
unusable-circid tracking.
2014-06-11 11:57:56 -04:00
Nick Mathewson
6557e61295 Replace last_added_nonpadding with last_had_circuits
The point of the "idle timeout" for connections is to kill the
connection a while after it has no more circuits.  But using "last
added a non-padding cell" as a proxy for that is wrong, since if the
last circuit is closed from the other side of the connection, we
will not have sent anything on that connection since well before the
last circuit closed.

This is part of fixing 6799.

When applied to 0.2.5, it is also a fix for 12023.
2014-06-11 11:27:04 -04:00
Nick Mathewson
463f6628d3 Give each or_connection_t a slightly randomized idle_timeout
Instead of killing an or_connection_t that has had no circuits for
the last 3 minutes, give every or_connection_t a randomized timeout,
so that an observer can't so easily infer from the connection close
time the time at which its last circuit closed.

Also, increase the base timeout for canonical connections from 3
minutes to 15 minutes.

Fix for ticket 6799.
2014-06-11 11:27:04 -04:00
Yawning Angel
1210bdf146 Log the correct proxy type on failure.
get_proxy_addrport fills in proxy_type with the correct value, so there
is no point in logging something that's a "best guess" based off the
config.
2014-05-21 08:14:39 +00:00
Yawning Angel
cd56b1a86e Remove get_bridge_pt_addrport().
The code was not disambiguating ClientTransportPlugin configured and
not used, and ClientTransportPlugin configured, but in a failed state.

The right thing to do is to undo moving the get_transport_by_addrport()
call back into get_proxy_addrport(), and remove and explicit check for
using a Bridge since by the time the check is made, if a Bridge is
being used, it is PT/proxy-less.
2014-05-21 08:14:39 +00:00
Yawning Angel
41d2b4d3af Allow ClientTransportPlugins to use proxies
This change allows using Socks4Proxy, Socks5Proxy and HTTPSProxy with
ClientTransportPlugins via the TOR_PT_PROXY extension to the
pluggable transport specification.

This fixes bug #8402.
2014-05-21 08:14:38 +00:00
Nick Mathewson
9b4ac986cb Use tor_getpw{nam,uid} wrappers to fix bug 11946
When running with User set, we frequently try to look up our
information in the user database (e.g., /etc/passwd).  The seccomp2
sandbox setup doesn't let us open /etc/passwd, and probably
shouldn't.

To fix this, we have a pair of wrappers for getpwnam and getpwuid.
When a real call to getpwnam or getpwuid fails, they fall back to a
cached value, if the uid/gid matches.

(Granting access to /etc/passwd isn't possible with the way we
handle opening files through the sandbox.  It's not desirable either.)
2014-05-14 13:53:14 -04:00
Nick Mathewson
d3c05a79f0 Merge branch 'scanbuild_fixes' 2014-04-25 01:24:39 -04:00
Nick Mathewson
895b6789e8 scan-build: get_proxy_addrport should always set its outputs
When get_proxy_addrport returned PROXY_NONE, it would leave
addr/port unset. This is inconsistent, and could (if we used the
function in a stupid way) lead to undefined behavior. Bugfix on
5b050a9b0, though I don't think it affects tor-as-it-is.
2014-04-18 20:41:40 -04:00
Nick Mathewson
bd169aa9a5 Merge remote-tracking branch 'public/bug11553_024' into bug11553_025
Conflicts:
	src/or/channel.h
2014-04-18 13:00:45 -04:00
Nick Mathewson
0d75344b0e Switch to random allocation on circuitIDs.
Fixes a possible root cause of 11553 by only making 64 attempts at
most to pick a circuitID.  Previously, we would test every possible
circuit ID until we found one or ran out.

This algorithm succeeds probabilistically. As the comment says:

  This potentially causes us to give up early if our circuit ID
  space is nearly full.  If we have N circuit IDs in use, then we
  will reject a new circuit with probability (N / max_range) ^
  MAX_CIRCID_ATTEMPTS.  This means that in practice, a few percent
  of our circuit ID capacity will go unused.

  The alternative here, though, is to do a linear search over the
  whole circuit ID space every time we extend a circuit, which is
  not so great either.

This makes new vs old clients distinguishable, so we should try to
batch it with other patches that do that, like 11438.
2014-04-18 12:58:58 -04:00
Nick Mathewson
c856193199 Merge remote-tracking branch 'andrea/bug11304' 2014-04-16 23:13:30 -04:00
Andrea Shepard
65a0f895c7 Check for orconns and use connection_or_close_for_error() when appropriate in connection_handle_write_impl() 2014-04-15 23:03:16 -07:00
Andrea Shepard
6ee9138576 Call connection_or_close_for_error() properly if write_to_buf() ever fails on an orconn 2014-04-15 21:25:49 -07:00
Andrea Shepard
a5544e589d Close orconns correctly through channels when setting DisableNetwork to 1 2014-04-15 20:19:39 -07:00
Nick Mathewson
d290e36576 Fix make_socket_reusable() on windows. Bug not in any released Tor 2014-04-02 21:11:45 -04:00
Nick Mathewson
60abc4804f Don't warn when setsockopt(SO_REUSEABLE) on accept()ed socket says EINVAL
This should fix bug10081.  I believe this bug pertains to OSX
behavior, not any Tor behavior change.
2014-03-27 13:55:18 -04:00
Karsten Loesing
7450403410 Take out remaining V1 directory code. 2014-03-18 10:40:10 +01:00
Nick Mathewson
cce06b649e Merge remote-tracking branch 'asn/bug11069_take2' 2014-03-11 11:04:47 -04:00
George Kadianakis
1c475eb018 Throw control port warning if we failed to connect to all our bridges. 2014-03-10 22:52:07 +00:00
Roger Dingledine
bd49653f8e trivial whitespace fixes 2014-03-03 06:53:08 -05:00
Nick Mathewson
b3a6907493 Remove a bunch of functions that were never called. 2014-02-15 15:33:34 -05:00
Nick Mathewson
372adfa09a Merge remote-tracking branch 'origin/maint-0.2.4' 2014-02-07 10:38:24 -05:00
Andrea Shepard
707c1e2e26 NULL out conns on tlschans when freeing in case channel_run_cleanup() is late; fixes bug 9602 2014-02-06 14:47:34 -08:00
Nick Mathewson
b4e8d8dc0e Merge remote-tracking branch 'public/bug9716_024' into maint-0.2.4 2014-02-06 16:29:08 -05:00
Nick Mathewson
5991f9a156 TransProxyType replaces TransTPROXY option
I'm making this change now since ipfw will want its own option too,
and proliferating options here isn't sensible.

(See #10582 and #10267)
2014-02-03 13:56:19 -05:00
Nick Mathewson
fd8947afc2 Move the friendly warning about TPROXY and root to EPERM time
I'm doing this because:
   * User doesn't mean you're running as root, and running as root
     doesn't mean you've set User.
   * It's possible that the user has done some other
     capability-based hack to retain the necessary privileges.
2014-02-02 15:45:00 -05:00
Nick Mathewson
09ccc4c4a3 Add support for TPROXY via new TransTPRoxy option
Based on patch from "thomo" at #10582.
2014-01-31 12:59:35 -05:00
Nick Mathewson
8f793c38fb Move other #9731 check to start of its functions
At arma's suggestion.  Looks like I missed this one.
2013-10-31 14:29:18 -04:00
Nick Mathewson
96f92f2062 Move #9731 checks to start of their functions
At arma's suggestion.
2013-10-31 14:10:23 -04:00
Nick Mathewson
702c0502cf Merge remote-tracking branch 'public/bug9731' 2013-10-31 14:09:18 -04:00
Karsten Loesing
2e0fad542c Merge branch 'morestats4' into morestats5
Conflicts:
	doc/tor.1.txt
	src/or/config.c
	src/or/connection.h
	src/or/control.c
	src/or/control.h
	src/or/or.h
	src/or/relay.c
	src/or/relay.h
	src/test/test.c
2013-10-28 12:09:42 +01:00
Nick Mathewson
4b8282e50c Log the origin address of controller connections
Resolves 9698; patch from "sigpipe".
2013-10-25 11:52:45 -04:00
Nick Mathewson
67bb1a44bd Merge remote-tracking branch 'public/bug9716_024' 2013-09-19 10:50:34 -04:00
Nick Mathewson
c5532889a8 Don't apply read/write buckets to cpuworker connections
Fixes bug 9731
2013-09-13 13:39:18 -04:00
Nick Mathewson
9205552daa Don't apply read/write buckets to non-limited connections
Fixes bug 9731
2013-09-13 13:37:53 -04:00
Karsten Loesing
d5f0d792dd Pass around const struct timeval * instead of struct timeval.
Suggested by nickm.
2013-09-12 10:10:38 +02:00
Nick Mathewson
a2754d418d Try using INT_MAX, not SOMAXCONN, to set listen() backlog.
Fall back to SOMAXCONN if INT_MAX doesn't work.

We'd like to do this because the actual maximum is overrideable by the
kernel, and the value in the header file might not be right at all.
All implementations I can find out about claim that this is supported.

Fix for 9716; bugfix on every Tor.
2013-09-11 13:30:45 -04:00
Nick Mathewson
74262f1571 Merge branch 'bug5040_4773_rebase_3' 2013-08-15 12:04:56 -04:00
Nick Mathewson
03e3881043 Tests for connection_write_ext_or_command. 2013-08-15 12:03:36 -04:00
Nick Mathewson
c342ea9879 Unit tests for ext_or_id_map. 2013-08-15 12:03:36 -04:00
Peter Retzlaff
ebd4ab1506 Prepare patch for ticket 5129 for merging.
- Preserve old eventdns code.
- Add function to close sockets cross-platform, without accounting.
- Add changes/ file.
2013-08-02 09:35:24 -04:00
George Kadianakis
210210f219 Make the Extended ORPort understand the TRANSPORT command. 2013-07-18 14:59:56 -04:00
George Kadianakis
c46f1b810d More Extended ORPort code improvements.
* Change name of init_ext_or_auth_cookie_authentication().
* Add a small comment.
2013-07-18 14:59:56 -04:00
George Kadianakis
d8f74cc439 Move Extended ORPort code to its own module.
Move the code from the connection_or module to ext_orport.

This commit only moves code: it shouldn't modify anything.
2013-07-18 14:59:56 -04:00
George Kadianakis
e2e0d09dab Various Extended ORPort code improvements.
* Add documentation.
* Free ext_or_auth_correct_client_hash.
* Use VPORT(ExtORPort) instead of V(ExtORPOrt).
  See dfe03d36c8 for details.
2013-07-18 14:59:56 -04:00
George Kadianakis
93b9f85d41 Prepare codebase for the implementation of Extended ORPort auth. 2013-07-18 14:59:56 -04:00
George Kadianakis
d303228eca Create the Extended ORPort authentication cookie file. 2013-07-18 14:59:55 -04:00
Nick Mathewson
8bf0382b22 Skeleton ExtORPort implementation. Needs testing, documentation.
Does not implement TransportControlPort yet.
2013-07-18 14:59:55 -04:00
Peter Palfrader
2cb59be999 Fix two pre-coffee typos 2013-07-15 09:43:37 -04:00
Peter Palfrader
783c52b6df Reject relative control socket paths and emit a warning.
Previously we would accept relative paths, but only if they contained a
slash somewhere (not at the end).

Otherwise we would silently not work.  Closes: #9258.  Bugfix on
0.2.3.16-alpha.
2013-07-15 09:04:17 -04:00
Karsten Loesing
e39292f21d Test functions used for TB_EMPTY and CELL_STATS events. 2013-05-31 16:12:28 +02:00
Karsten Loesing
b33b366a7f Tweak CIRC_BW event based on comments by nickm.
- Rename n_read and n_written in origin_circuit_t to make it clear that
  these are only used for CIRC_BW events.
- Extract new code in control_update_global_event_mask to new
  clear_circ_bw_fields function.
2013-05-25 19:51:38 +02:00
Karsten Loesing
ef67077fba Tweak TB_EMPTY event based on comments by nickm.
- Avoid control_event_refill_global function with 13 arguments and
  increase code reuse factor by moving more code from control.c to
  connection.c.
- Avoid an unsafe uint32_t -> int cast.
- Add TestingEnableTbEmptyEvent option.
- Prepare functions for testing.
- Rename a few functions and improve documentation.
2013-05-25 19:51:38 +02:00
Karsten Loesing
2f893624ab Tweak CONN_BW event based on comments by nickm.
- Rename read/write counters in connection_t to make it clear that these
  are only used for CONN_BW events.
- Add TestingEnableConnBwEvent option.
2013-05-25 19:51:38 +02:00
Nick Mathewson
b4b0063e48 Tweak fix for #8789 a bit; avoid double-close and add changes file 2013-05-24 12:23:21 -04:00
Arlo Breault
d25e77f2c3 Close socket at err target.
In connection_listener_new().
See #8789.
2013-05-24 12:05:57 -04:00
Karsten Loesing
2925e2fe78 Add new CIRC_BW event. 2013-05-16 14:18:13 +02:00
Karsten Loesing
dd5ce2157d Add new TB_EMPTY event.
Jointly authored with Rob Jansen <jansen@cs.umn.edu>.
2013-05-16 14:18:08 +02:00
Karsten Loesing
8d1f78c556 Add new CONN_BW event.
Jointly authored with Rob Jansen <jansen@cs.umn.edu>.
2013-05-16 13:48:35 +02:00