Commit Graph

13098 Commits

Author SHA1 Message Date
Robert Ransom
2ad18ae736 Allow rend_client_send_introduction to fail transiently
i.e. without closing the AP connection.
2011-04-28 18:14:50 +02:00
Robert Ransom
440e48ddf2 Forget all rendezvous client state on SIGNAL NEWNYM 2011-04-28 18:14:50 +02:00
Nick Mathewson
0130e7c9d2 Merge remote-tracking branch 'origin/maint-0.2.1' into maint-0.2.2
Conflicts:
	src/common/torint.h
2011-04-27 22:14:28 -04:00
Nick Mathewson
7f85509a59 I guess that had a bug number: add it to the changes file. 2011-04-27 17:26:28 -04:00
Nick Mathewson
34510f9278 Fix clear_trackhostexits_mapping() to actually work as advertised
Previously, it would remove every trackhostexits-derived mapping
*from* xyz.<exitname>.exit; it was supposed to remove every
trackhostexits-derived mapping *to* xyz.<exitname>.exit.

Bugfix on 0.2.0.20-rc: fixes an XXX020 added while staring at bug-1090
issues.
2011-04-27 17:23:05 -04:00
Roger Dingledine
cba1d29b7f make make check-spaces happier.
(still not happy.)
2011-04-27 15:18:34 -04:00
cypherpunks
247cbab6c8 Fix double-free bug in microdesc parser 2011-04-27 15:15:32 -04:00
Roger Dingledine
0f48e8fa9a minor tweaks to manpage 2011-04-27 14:04:50 -04:00
Nick Mathewson
99621bc5a6 Merge remote-tracking branch 'arma/bug1090-part1-squashed' into bug1090-part1-squashed 2011-04-27 13:45:28 -04:00
Nick Mathewson
8f6cb933b6 Merge branch 'bug1090-manpage-squashed' into bug1090-part1-squashed 2011-04-27 13:44:55 -04:00
Roger Dingledine
0c40dda3ad explain an argument in a function comment 2011-04-27 13:43:11 -04:00
Nick Mathewson
2ac768e89f Revise the manpage to contain the actual intended *Nodes behavior
This is a squashed version of my former desired_nodes_behavior branch
that we used to specify the intended results wrt bug 1090.
2011-04-27 13:39:22 -04:00
Nick Mathewson
c49f660c1a Add a big changelog entry for bug 1090 fixes 2011-04-27 13:33:38 -04:00
Nick Mathewson
748350ace1 Instead of checking whether we have unremoved intro points, check for usable ones 2011-04-27 00:01:41 -04:00
Roger Dingledine
b8b557dcb2 better user-facing warnings for unexpected last hops
these still aren't perfect, but we won't know how to correct them until
we start experiencing surprised users.
2011-04-27 00:01:41 -04:00
Roger Dingledine
f7a5bc16d6 warn if we launch too many circuits for a given stream 2011-04-27 00:01:41 -04:00
Roger Dingledine
f962dda8c1 revert most of ef81649d2f
Now we believe it to be the case that we never build a circuit for our
stream that has an unsuitable exit, so we'll never need to use such
a circuit. The risk is that we have some code that builds the circuit,
but now we refuse to use it, meaning we just build a bazillion circuits
and ignore them all.
2011-04-27 00:01:41 -04:00
Nick Mathewson
8ee92f28e0 Add a circuit_purpose_to_string() function, and use it
We had a circuit_purpose_to_controller_string() function, but it was
pretty coarse-grained and didn't try to be human-readable.
2011-04-27 00:01:35 -04:00
Sebastian Hahn
e03e90bf59 Fix a check-spaces complaint 2011-04-26 23:55:23 -04:00
Sebastian Hahn
92ec36a061 Explain the "using anyway" log message better
Also add a request to report any cases where we are not honoring
StrictNodes to the developers: These should now all be bugs.
2011-04-26 23:55:23 -04:00
Sebastian Hahn
8e2904e269 Fix a log msg 2011-04-26 23:55:22 -04:00
Nick Mathewson
79a3b3cd37 Check transition of circuit purpose from INTRO->GENERAL if nodes are constrained
This looked at first like another fun way around our node selection
logic: if we had introduction circuits, and we wound up building too
many, we would turn extras into general-purpose circuits.  But when we
did so, we wouldn't necessarily check whether the general-purpose
circuits conformed to our node constraints.  For example, the last
node could totally be in ExcludedExitNodes and we wouldn't have cared...

...except that the circuit should already be internal, so it won't get user
streams attached to it, so the transition should generally be allowed.
Add an assert to make sure we're right about this, and have it not
check whether ExitNodes is set, since that's irrelevant to internal
circuits.
2011-04-26 23:54:18 -04:00
Nick Mathewson
6afad6b691 When cannibalizing a circuit, make sure it has no ExcludeNodes on it
This could happen if StrictNodes was 0 and we were forced to pick an
excluded node as the last hop of the circuit.
2011-04-26 23:54:17 -04:00
Nick Mathewson
80adb3de50 When there is a transition in permitted nodes, apply it to trackexithosts map
IOW, if we were using TrackExitHosts, and we added an excluded node or
removed a node from exitnodes, we wouldn't actually remove the mapping
that points us at the new node.

Also, note with an XXX022 comment a place that I think we are looking
at the wrong string.
2011-04-26 23:54:17 -04:00
Nick Mathewson
128582cc1f Simplify calls to routerset_equal
The routerset_equal function explicitly handles NULL inputs, so
there's no need to check inputs for NULL before calling it.

Also fix a bug in routerset_equal where a non-NULL routerset with no
entries didn't get counted as equal to a NULL routerset.  This was
untriggerable, I think, but potentially annoying down the road.
2011-04-26 23:54:17 -04:00
Nick Mathewson
84f0e87c6a If we have chosen an exit that shares a family with all bridges, fail the circuit
We could probably do something smarter here, but the situation is
unusual enough that it's okay to just fail the circuit.
2011-04-26 23:54:17 -04:00
Nick Mathewson
b59a289365 Do not try to download descriptors for bridges in ExcludeNodes. 2011-04-26 23:54:16 -04:00
Nick Mathewson
ad78bafb71 Correct the behavior of .exit with ExcludeNodes, StrictNodes, etc.
ExcludeExitNodes foo now means that foo.exit doesn't work.  If
StrictNodes is set, then ExcludeNodes foo also overrides foo.exit.

foo.exit , however, still works even if foo is not listed in ExitNodes.
2011-04-26 23:54:16 -04:00
Nick Mathewson
ed7c267743 Note another place that we need to fix a 1090 issue. 2011-04-26 23:54:16 -04:00
Nick Mathewson
4851de554d Do not automatically ignore Fast/Stable for exits when ExitNodes is set
This once maybe made sense when ExitNodes meant "Here are 3 exits;
use them all", but now it more typically means "Here are 3
countries; exit from there."  Using non-Fast/Stable exits created a
potential partitioning opportunity and an annoying stability
problem.

(Don't worry about the case where all of our ExitNodes are non-Fast
or non-Stable: we handle that later in the function by retrying with
need_capacity and need_uptime set to 0.)
2011-04-26 23:54:16 -04:00
Nick Mathewson
e4689d8402 Note a slightly less likely way to violate ExcludeNodes 2011-04-26 23:54:16 -04:00
Nick Mathewson
db2fd28308 Note that circuit purpose changing can violate ExcludeNodes 2011-04-26 23:54:16 -04:00
Nick Mathewson
ca74badbe9 If we're excluded, and StrictNodes is set, do not do self-tests. 2011-04-26 23:54:15 -04:00
Nick Mathewson
affdec8d04 Add an XXX022-1090 to note consider_exit_fmily b0rkenness 2011-04-26 23:54:15 -04:00
Roger Dingledine
5710ea6475 three more cases where maybe we want to exclude 2011-04-26 23:54:15 -04:00
Roger Dingledine
9f47cfc21a make formal a constraint that's been true a while now 2011-04-26 23:54:15 -04:00
Roger Dingledine
2b5c39211c refuse moria1.exit if moria1 is excluded
add a note reminding us to do this for foo.moria1.exit if we decide to.
2011-04-26 23:54:15 -04:00
Roger Dingledine
bcea155ce0 note another case where strictnodes is considered for exits 2011-04-26 23:54:14 -04:00
Roger Dingledine
0ad3836f73 If ExitNodes and Exclude{Exit}Nodes overlap, obey Exclude{Exit}Nodes.
Also, ExitNodes are always strict.
2011-04-26 23:54:14 -04:00
Roger Dingledine
5d12495d98 the new entrynodes behavior is always strict 2011-04-26 23:54:13 -04:00
Roger Dingledine
719b5b87de don't exit enclave to excluded relays 2011-04-26 23:54:13 -04:00
Roger Dingledine
7e2e8074d5 slight tweak on circuit_conforms_to_options
this function really needs to get a total rewrite (or die)

For now, use #if 0 to disable it.
2011-04-26 23:54:04 -04:00
Roger Dingledine
4906188b62 handle excludenodes for dir fetch/post
If we're picking a random directory node, never pick an excluded one.
But if we've chosen a specific one (or all), allow it unless strictnodes
is set (in which case warn so the user knows it's their fault).

When warning that we won't connect to a strictly excluded node,
log what it was we were trying to do at that node.

When ExcludeNodes is set but StrictNodes is not set, we only use
non-excluded nodes if we can, but fall back to using excluded nodes
if none of those nodes is usable.
2011-04-26 23:53:50 -04:00
Roger Dingledine
ad3da53536 If EntryNodes and ExcludeNodes overlap, obey ExcludeNodes. 2011-04-26 23:53:49 -04:00
Roger Dingledine
82178a81f6 refuse excluded hidserv nodes if strictnodes
Make hidden services more flaky for people who set both ExcludeNodes
and StrictNodes. Not recommended, especially for hidden service operators.
2011-04-26 23:53:20 -04:00
Nick Mathewson
f810a1afe9 Expose a new process_signal(uintptr_t), not signal_callback()
This is a tweak to the bug2917 fix.  Basically, if we want to simulate
a signal arriving in the controller, we shouldn't have to pretend that
we're Libevent, or depend on how Tor sets up its Libevent callbacks.
2011-04-26 15:20:08 -04:00
Nick Mathewson
a7a906603e Merge remote-tracking branch 'sebastian/bug2917' into maint-0.2.2 2011-04-26 15:17:03 -04:00
Nick Mathewson
b75d1daf40 Merge remote-tracking branch 'public/bug2332' into maint-0.2.2 2011-04-26 13:10:54 -04:00
Nick Mathewson
43ffd023e9 Make SIZE_T_CEILING unsigned; add a signed SSIZE_T_CEILING
None of the comparisons were _broken_ previously, but avoiding
signed/unsigned comparisons makes everybody happier.

Fixes bug2475.
2011-04-26 13:03:58 -04:00
Nick Mathewson
bb6d45af1f Downgrade notice to info when downloading a cert. 2011-04-26 12:47:09 -04:00