Commit Graph

14630 Commits

Author SHA1 Message Date
Robert Ransom
00885652db Allow intro points to expire somewhat gracefully
The Right Way to expire an intro point is to establish a new one to
replace it, publish a new descriptor that doesn't list any expiring intro
points, and *then*, once our upload attempts for the new descriptor have
ended (whether in success or failure), close the expiring intro points.

Unfortunately, we can't find out when the new descriptor has actually been
uploaded, so we'll have to settle for a five-minute timer.

There should be no significant behaviour changes due to this commit (only
a log-message change or two), despite the rather massive overhaul, so this
commit doesn't include a changes/ file.  (The commit that teaches
intro_point_should_expire_now to return non-zero gets a changes/ file,
though.)
2011-10-30 02:17:10 -07:00
Robert Ransom
ae9a831294 Use SMARTLIST_FOREACH_BEGIN and _END, not a for loop 2011-10-30 02:17:07 -07:00
Robert Ransom
e46d56a9b4 Correct bogus comments
The behaviour of rend_services_introduce here is likely as bogus as the
comments were.
2011-10-30 02:17:06 -07:00
Robert Ransom
290b4dc3d6 Use a more meaningful variable name 2011-10-30 02:17:06 -07:00
Robert Ransom
68331cbd81 Determine whether an intro point was in the last HS desc in a sane way 2011-10-30 02:17:05 -07:00
Robert Ransom
3f6a2d3e2a Record which intro points were listed in the last HS desc 2011-10-30 02:17:05 -07:00
Robert Ransom
6f035cb2b4 Record the number of INTRODUCE2 cells each intro point has received 2011-10-30 02:17:04 -07:00
Robert Ransom
6b26999146 Correct a log message 2011-10-30 02:17:01 -07:00
Robert Ransom
c60148c7f5 Record the time at which each intro point was first published 2011-10-30 02:17:00 -07:00
Robert Ransom
6f220de7b9 Improve a comment 2011-10-30 02:17:00 -07:00
Robert Ransom
190aac0eab Allow different HSes to maintain different numbers of intro points 2011-10-30 02:16:59 -07:00
Roger Dingledine
46296b3fa4 there will be a new alpha release today (oct 30) 2011-10-30 01:58:05 -04:00
Roger Dingledine
eeb6588389 bridges send netinfo cells like clients on outgoing conns
fixes bug 4348
2011-10-29 21:43:23 -04:00
Roger Dingledine
c03b4fa826 amend the changelog entry for 4299
it doesn't have to do with small cells. rather, it has to do with
using cells to decide whether to mark the connection for close.
2011-10-29 16:24:59 -04:00
Roger Dingledine
cbdefc70fc fold in changes entries 2011-10-29 15:10:37 -04:00
Nick Mathewson
6be596e2dc Merge remote-tracking branch 'origin/maint-0.2.2'
Conflicts:
	src/or/config.c
2011-10-29 09:22:58 -04:00
Sebastian Hahn
c5c5f859d3 Add a changes file for the 4340 fix 2011-10-29 11:21:37 +02:00
Sebastian Hahn
dfdb4fde1a Disable stats requiring geoip info if we have none
In other parts of the code we will otherwise attempt to collect these
statistics, and that will lead to crashes.
2011-10-29 11:20:02 +02:00
Nick Mathewson
4dd8d811d6 Merge branch 'bug4343' 2011-10-28 18:05:25 -04:00
Nick Mathewson
00a0de8508 Bump version to 0.2.3.6-alpha-dev 2011-10-28 18:05:17 -04:00
Nick Mathewson
212c3acd42 Merge remote-tracking branch 'origin/maint-0.2.2'
Conflicts:
	src/or/command.c
2011-10-28 18:02:57 -04:00
Nick Mathewson
cf8bffa359 Discard all cells on a marked connection
Fix for bug 4299
2011-10-28 17:04:15 -04:00
Nick Mathewson
2018f86e0c "Authetnicate" is not the usual spelling 2011-10-28 16:41:04 -04:00
Nick Mathewson
c2a098e980 Fix a double-free that would occur on an invalid cert in a CERTS cell
We would stash the certs in the handshake state before checking them
for validity... and then if they turned out to be invalid, we'd give
an error and free them.  Then, later, we'd free them again when we
tore down the connection.

Fixes bug 4343; fix on 0.2.3.6-alpha.
2011-10-28 16:38:56 -04:00
Nick Mathewson
7a8960cf1b Fix a memory-poisoning memset in tortls.c 2011-10-28 16:37:42 -04:00
Nick Mathewson
a2517fa77c Merge remote-tracking branch 'origin/maint-0.2.2' 2011-10-28 09:53:23 -04:00
Sebastian Hahn
f6b1dc2805 Fix typo, spotted by tmpname0901. Thanks! 2011-10-28 09:52:24 -04:00
Roger Dingledine
ba91272ded Merge branch 'maint-0.2.2' 2011-10-27 20:53:17 -04:00
Roger Dingledine
48bafb47ac bump maint-0.2.2 to 0.2.2.34-dev 2011-10-27 20:50:48 -04:00
Roger Dingledine
6258013e41 find all those stanzas in master too 2011-10-27 20:15:50 -04:00
Roger Dingledine
425e4236c6 Merge branch 'maint-0.2.2' 2011-10-27 20:04:51 -04:00
Roger Dingledine
9c2ca5b52f Merge branch 'maint-0.2.1' into maint-0.2.2 2011-10-27 20:03:21 -04:00
Roger Dingledine
5d3095152e bump maint to 0.2.1.31 2011-10-27 20:01:58 -04:00
Roger Dingledine
47dff61061 a little blurb for 0.2.3.6-alpha 2011-10-26 20:54:15 -04:00
Roger Dingledine
821f59b4df forward-port the stable releases notes 2011-10-26 20:41:51 -04:00
Roger Dingledine
1a160ae523 bump to 0.2.3.6-alpha 2011-10-26 20:33:23 -04:00
Roger Dingledine
0eaebebffa fold in changes entries 2011-10-26 20:31:49 -04:00
Sebastian Hahn
2dec6597af Merge branch 'maint-0.2.2_secfix' into master_secfix
Conflicts:
	src/common/tortls.c
	src/or/connection_or.c
	src/or/dirserv.c
	src/or/or.h
2011-10-27 00:38:45 +02:00
Robert Ransom
4684ced1b3 Add option to give guard flag to relays without the CVE-2011-2768 fix
This way, all of the DA operators can upgrade immediately, without nuking
every client's set of entry guards as soon as a majority of them upgrade.

Until enough guards have upgraded, a majority of dirauths should set this
config option so that there are still enough guards in the network. After
a few days pass, all dirauths should use the default.
2011-10-26 23:56:47 +02:00
Robert Ransom
00fffbc1a1 Don't give the Guard flag to relays without the CVE-2011-2768 fix 2011-10-26 23:42:39 +02:00
Robert Ransom
4d0f152aad Make tor_version_same_series non-static 2011-10-26 23:36:30 +02:00
Sebastian Hahn
df05e5ef4d Merge branch 'maint-0.2.1_secfix' into maint-0.2.2_secfix
Conflicts:
	src/or/connection_or.c
2011-10-26 23:30:27 +02:00
Robert Ransom
a74e7fd40f Reject create cells on outgoing OR connections from bridges 2011-10-26 23:21:14 +02:00
Robert Ransom
c05bb53508 Mark which OR connections are outgoing 2011-10-26 23:21:11 +02:00
Robert Ransom
af12c39d6d Don't use any OR connection which sent us a CREATE_FAST cell for an EXTEND
Fix suggested by Nick Mathewson.
2011-10-26 23:20:56 +02:00
Nick Mathewson
638fdedcf1 Don't send a certificate chain on outgoing TLS connections from non-relays 2011-10-26 23:20:56 +02:00
Roger Dingledine
2c4e89b867 Merge branch 'maint-0.2.2' 2011-10-26 17:15:09 -04:00
Roger Dingledine
f85c56182f Merge branch 'maint-0.2.1' into maint-0.2.2 2011-10-26 17:14:39 -04:00
Nick Mathewson
a166f10414 Remove the -F option from tor-resolve.
It used to mean "Force": it would tell tor-resolve to ask tor to
resolve an address even if it ended with .onion.  But when
AutomapHostsOnResolve was added, automatically refusing to resolve
.onion hosts stopped making sense.  So in 0.2.1.16-rc (commit
298dc95dfd), we made tor-resolve happy to resolve anything.

The -F option stayed in, though, even though it didn't do anything.
Oddly, it never got documented.

Found while fixing GCC 4.6 "set, unused variable" warnings.
2011-10-26 17:13:04 -04:00
Roger Dingledine
a68867b150 manually backport a5232e0c4c 2011-10-26 17:11:52 -04:00