]> git.ipfire.org Git - thirdparty/tor.git/log
thirdparty/tor.git
16 years agotor-resolve: Don't automatically refuse .onion addresses.
Nick Mathewson [Thu, 18 Jun 2009 15:08:10 +0000 (11:08 -0400)] 
tor-resolve: Don't automatically refuse .onion addresses.

If the Tor is running with AutomapHostsOnResolve set, it _is_
reasonable to do a DNS lookup on a .onion address.  So instead we make
tor-resolve willing to try to resolve anything.  Only if Tor refuses
to resolve it do we suggest to the user that resolving a .onion
address may not work.

Fix for bug 1005.

16 years agoBackport fix for bug 997.
Karsten Loesing [Tue, 16 Jun 2009 20:07:16 +0000 (22:07 +0200)] 
Backport fix for bug 997.

Backporting 6a32beb and ca8708a.

16 years agoRevise earlier check for correct IPv4 addr length to check for ==4.
Nick Mathewson [Tue, 16 Jun 2009 20:12:06 +0000 (16:12 -0400)] 
Revise earlier check for correct IPv4 addr length to check for ==4.

We need this to match the check in connection_ap_handshake_socks_resolved().

Found by optimist.

16 years agoCheck answer_len in the remap_addr case of process_relay_cell_not_open.
Roger Dingledine [Fri, 12 Jun 2009 15:18:02 +0000 (11:18 -0400)] 
Check answer_len in the remap_addr case of process_relay_cell_not_open.

Fix an edge case where a malicious exit relay could convince a
controller that the client's DNS question resolves to an internal IP
address. Bug found and fixed by "optimist"; bugfix on 0.1.2.8-beta.

16 years agoupdate requirements to openssl 0.9.7
Andrew Lewman [Mon, 8 Jun 2009 14:27:10 +0000 (10:27 -0400)] 
update requirements to openssl 0.9.7

16 years agoMissing changelog entry about geoip
Nick Mathewson [Mon, 8 Jun 2009 04:00:54 +0000 (00:00 -0400)] 
Missing changelog entry about geoip

17 years agoMove and fix a changelog entry. Noticed by optimist.
Nick Mathewson [Sun, 7 Jun 2009 04:03:27 +0000 (00:03 -0400)] 
Move and fix a changelog entry.  Noticed by optimist.

17 years agoAvoid a memory corruption problem related to "private" in DirPolicy.
Nick Mathewson [Fri, 5 Jun 2009 23:41:40 +0000 (19:41 -0400)] 
Avoid a memory corruption problem related to "private" in DirPolicy.

This is a posible fix for bug 996.

17 years agoUpdate the rest of the geoip file.
Karsten Loesing [Wed, 3 Jun 2009 22:47:58 +0000 (00:47 +0200)] 
Update the rest of the geoip file.

17 years agoUpdate the geoip file
Nick Mathewson [Wed, 3 Jun 2009 21:35:41 +0000 (17:35 -0400)] 
Update the geoip file

17 years agoFix gprof bottlenecks on exit nodes found by Jacob.
Nick Mathewson [Wed, 3 Jun 2009 17:52:03 +0000 (13:52 -0400)] 
Fix gprof bottlenecks on exit nodes found by Jacob.

Apparently all the stuff that does a linear scan over all the DNS
cache entries can get really expensive when your DNS cache is very
large.  It's hard to say how much this will help performance, since
gprof doesn't count time spent in OpenSSL or zlib, but I'd guess 10%.

Also, this patch removes calls to assert_connection_ok() from inside
the read and write callbacks, which are similarly unneeded, and a
little costlier than I'm happy with.

This is probably worth backporting to 0.2.0.

17 years agoDo not report a node as a "chosen exit" when it is not in fact an exit.
Nick Mathewson [Wed, 27 May 2009 18:07:41 +0000 (14:07 -0400)] 
Do not report a node as a "chosen exit" when it is not in fact an exit.

Provide a useful warning when launch_circuit tries to make us use a
node we don't want to use.  Just give an info message when this is a
normal and okay situation.  Fix for logging issues in bug 984.

17 years agoMake the second argument to routerset_contains_extendinfo const
Nick Mathewson [Wed, 27 May 2009 18:45:44 +0000 (14:45 -0400)] 
Make the second argument to routerset_contains_extendinfo const

17 years agoDon't attempt to log messages to a controller from a worker thread.
Nick Mathewson [Fri, 29 May 2009 14:18:50 +0000 (10:18 -0400)] 
Don't attempt to log messages to a controller from a worker thread.

This patch adds a function to determine whether we're in the main
thread, and changes control_event_logmsg() to return immediately if
we're in a subthread.  This is necessary because otherwise we will
call connection_write_to_buf, which modifies non-locked data
structures.

Bugfix on 0.2.0.x; fix for at least one of the things currently
called "bug 977".

17 years agoClean up a bit of C logic, and fix an erroneous warning.
Nick Mathewson [Sat, 30 May 2009 03:28:01 +0000 (23:28 -0400)] 
Clean up a bit of C logic, and fix an erroneous warning.

(Simplify "if (cond) return 1; return 0;" to "return cond;", and don't
give a warning when we start with accounting on but dirport off.)

17 years agoConsider *ListenAddress when warning about low ports and hibernation
Sebastian Hahn [Tue, 26 May 2009 10:41:45 +0000 (12:41 +0200)] 
Consider *ListenAddress when warning about low ports and hibernation

Tas (thanks!) noticed that when *ListenAddress is set, Tor would
still warn on startup when *Port is low and hibernation is active.
The patch parses all the *ListenAddress lines, and checks the
ports. Bugfix on 0.2.1.15-rc

17 years agoTwo more spelling fix fixes.
Nick Mathewson [Thu, 28 May 2009 16:26:17 +0000 (12:26 -0400)] 
Two more spelling fix fixes.

17 years agoFixes to spelling fixes. Thanks, Roger!
Nick Mathewson [Thu, 28 May 2009 15:54:56 +0000 (11:54 -0400)] 
Fixes to spelling fixes.  Thanks, Roger!

17 years agoFix unit tests that were broken after last fix of #932.
Karsten Loesing [Wed, 27 May 2009 21:47:59 +0000 (23:47 +0200)] 
Fix unit tests that were broken after last fix of #932.

With the last fix of task 932 (5f03d6c), client requests are only added to
the history when they happen after the start of the current history. This
conflicts with the unit tests that insert current requests first (defining
the start of the client request history) followed by requests in the past.
The fix is to insert requests in chronological order in the unit tests.

17 years agoSpell-check Tor.
Nick Mathewson [Wed, 27 May 2009 21:55:51 +0000 (17:55 -0400)] 
Spell-check Tor.

17 years agoSpelling fixes in comments and strings
Nick Mathewson [Wed, 27 May 2009 20:35:03 +0000 (16:35 -0400)] 
Spelling fixes in comments and strings

17 years agoChangelog entry for Karsten's bug-932 patch.
Nick Mathewson [Wed, 27 May 2009 18:11:38 +0000 (14:11 -0400)] 
Changelog entry for Karsten's bug-932 patch.

17 years agoImprove documentation for the last fix of bug 932.
Karsten Loesing [Tue, 26 May 2009 20:02:52 +0000 (22:02 +0200)] 
Improve documentation for the last fix of bug 932.

17 years agoFix bug 932 even more.
Karsten Loesing [Tue, 26 May 2009 19:41:42 +0000 (21:41 +0200)] 
Fix bug 932 even more.

Ignore connections two hours after switching from bridge to relay or back.

17 years agoAdd *.swp to .gitignore as vim's editor dropping.
Nick Mathewson [Wed, 27 May 2009 16:10:37 +0000 (12:10 -0400)] 
Add *.swp to .gitignore as vim's editor dropping.

17 years agoUse tor_socket_strerror in eventdns.c. Fixes bug 987.
Nick Mathewson [Mon, 25 May 2009 20:38:50 +0000 (16:38 -0400)] 
Use tor_socket_strerror in eventdns.c. Fixes bug 987.

17 years agoBump version to 0.2.1.15-rc-dev.
Nick Mathewson [Mon, 25 May 2009 20:32:47 +0000 (16:32 -0400)] 
Bump version to 0.2.1.15-rc-dev.

17 years agobump to 0.2.1.15-rc tor-0.2.1.15-rc
Roger Dingledine [Mon, 25 May 2009 17:23:08 +0000 (13:23 -0400)] 
bump to 0.2.1.15-rc

17 years agoadd the 0.2.1.14-rc blurb
Roger Dingledine [Mon, 25 May 2009 17:16:45 +0000 (13:16 -0400)] 
add the 0.2.1.14-rc blurb

17 years agoclean up the 0.2.1.15-rc changelog
Roger Dingledine [Mon, 25 May 2009 17:14:58 +0000 (13:14 -0400)] 
clean up the 0.2.1.15-rc changelog

17 years agoFinish up the 0.2.1 version of the bug 957 fix.
Nick Mathewson [Fri, 22 May 2009 15:47:09 +0000 (11:47 -0400)] 
Finish up the 0.2.1 version of the bug 957 fix.

Basically, all this means is downgrading our warning messages to debug
messages, since the bug workaround code here is adequate to stop the
bug.

17 years agofix a spelling mistake in config.c for "contries"
phobos [Fri, 22 May 2009 05:19:55 +0000 (01:19 -0400)] 
fix a spelling mistake in config.c for "contries"

17 years agoPossible fix for crash bug related to event timeouts. [Bug 957]
Nick Mathewson [Mon, 18 May 2009 20:12:39 +0000 (16:12 -0400)] 
Possible fix for crash bug related to event timeouts. [Bug 957]

If we ever add an event, then set it, then add it again, there will be
now two pointers to the event in the event base.  If we delete one and
free it, the first pointer will still be there, and possibly cause a
crash later.

This patch adds detection for this case to the code paths in
eventdns.c, and works around it.  If the warning message ever
displays, then a cleverer fix is in order.

{I am not too confident that this *is* the fix, since bug 957 is very
tricky.  If it is, it is a bugfix on 0.2.0.}

17 years agoFix a memory leak when v3 directory authorities load their keys
Roger Dingledine [Sun, 17 May 2009 05:22:29 +0000 (01:22 -0400)] 
Fix a memory leak when v3 directory authorities load their keys
and cert from disk. Bugfix on 0.2.0.1-alpha.

17 years agoFix valgrind error when marking a descriptor as never-downloadable.
Nick Mathewson [Sun, 17 May 2009 06:01:09 +0000 (02:01 -0400)] 
Fix valgrind error when marking a descriptor as never-downloadable.

When we got a descriptor that we (as an authority) rejected as totally
bad, we were freeing it, then using the digest in its RAM to look up its
download status.  Caught by arma with valgrind.  Bugfix on 0.2.1.9-alpha.

17 years agoStop using malloc_usable_size(): valgrind hates it.
Nick Mathewson [Sun, 17 May 2009 05:55:02 +0000 (01:55 -0400)] 
Stop using malloc_usable_size(): valgrind hates it.

17 years agominor tweak on changelog entries
Roger Dingledine [Sun, 17 May 2009 04:51:23 +0000 (00:51 -0400)] 
minor tweak on changelog entries

17 years agoFix an assertion-failure in memarea_alloc() on 64-bit platforms.
Nick Mathewson [Sun, 17 May 2009 03:57:30 +0000 (23:57 -0400)] 
Fix an assertion-failure in memarea_alloc() on 64-bit platforms.

The trick is that we should assert that our next_mem pointer has not
run off the end of the array _before_ we realign the pointer, since
doing that could take us over the end... but only if we're on a system
where malloc() gives us ram in increments smaller than sizeof(void*).

17 years agoFix misreporting of stream bandwidths.
Mike Perry [Thu, 14 May 2009 12:20:27 +0000 (05:20 -0700)] 
Fix misreporting of stream bandwidths.

17 years agoChangelog entries for bug932 fixes
Nick Mathewson [Wed, 13 May 2009 20:44:10 +0000 (16:44 -0400)] 
Changelog entries for bug932 fixes

17 years agoUse | with flags, not +.
Nick Mathewson [Wed, 13 May 2009 20:43:51 +0000 (16:43 -0400)] 
Use | with flags, not +.

17 years agoPrevent bridges from publishing router descriptors.
Karsten Loesing [Wed, 13 May 2009 16:14:29 +0000 (18:14 +0200)] 
Prevent bridges from publishing router descriptors.

Bridges are not supposed to publish router descriptors to the directory
authorities. It defeats the point of bridges when they are included in the
public relay directory.

This patch puts out a warning and exits when the node is configured as
a bridge and to publish v1, v2, or v3 descriptors at the same time.

Also fixes part of bug 932.

17 years agoWhen our bridge status changes, forget old geoip data.
Nick Mathewson [Tue, 12 May 2009 19:35:53 +0000 (15:35 -0400)] 
When our bridge status changes, forget old geoip data.

This fixes bug 932, where all of our usage totals for users when we
were a regular server would get included in the extrainfo document.

17 years agoUse a mutex to protect the count of open sockets.
Nick Mathewson [Tue, 12 May 2009 20:17:32 +0000 (16:17 -0400)] 
Use a mutex to protect the count of open sockets.

This matters because a cpuworker can close its socket when it
finishes.  Cpuworker typically runs in another thread, so without a
lock here, we can have a race condition and get confused about how
many sockets are open.  Possible fix for bug 939.

17 years agoWarn when hibernation and low-port on non-windows is configured
Sebastian Hahn [Tue, 12 May 2009 17:48:08 +0000 (19:48 +0200)] 
Warn when hibernation and low-port on non-windows is configured

This addresses the first part of bug 918. Users are now warned when
they try to use hibernation in combination with a port below 1024
when they're not on Windows. We don't want to die here, because
people might run Tor as root, use a capabilities system or some
other platform that will allow them to re-attach low ports.

Wording suggested by Marian

17 years agoBetter debugging output for bug 977 case.
Nick Mathewson [Tue, 12 May 2009 17:54:21 +0000 (13:54 -0400)] 
Better debugging output for bug 977 case.

(Don't crash immediately if we have leftover chunks to free after
freeing chunks in a buffer freelist; instead log a debugging message
that might help.)

17 years agoUpdate copyright to 2009.
Karsten Loesing [Sat, 2 May 2009 20:00:54 +0000 (22:00 +0200)] 
Update copyright to 2009.

17 years agoBump 0.2.1 branch version to 0.2.1.14-rc-dev
Nick Mathewson [Wed, 29 Apr 2009 18:16:15 +0000 (14:16 -0400)] 
Bump 0.2.1 branch version to 0.2.1.14-rc-dev

17 years agoAdd foo.o and foo~ files to gitignore.
Nick Mathewson [Wed, 29 Apr 2009 18:00:14 +0000 (14:00 -0400)] 
Add foo.o and foo~ files to gitignore.

17 years agoAdd an initial .gitignore to master
Nick Mathewson [Wed, 29 Apr 2009 13:19:18 +0000 (09:19 -0400)] 
Add an initial .gitignore to master

17 years agofix typo and naming convention
Roger Dingledine [Sun, 26 Apr 2009 03:59:35 +0000 (03:59 +0000)] 
fix typo and naming convention

svn:r19377

17 years agoInitial thoughts on a TODO.022.
Nick Mathewson [Fri, 24 Apr 2009 05:45:37 +0000 (05:45 +0000)] 
Initial thoughts on a TODO.022.

svn:r19369

17 years agoReturn -1 in the error case from read_bandwidth_usage.
Sebastian Hahn [Thu, 23 Apr 2009 22:04:10 +0000 (22:04 +0000)] 
Return -1 in the error case from read_bandwidth_usage.

svn:r19367

17 years agoUpdate the port knocking SPA document to have more details. Still needs a packet...
Jacob Appelbaum [Tue, 21 Apr 2009 07:55:07 +0000 (07:55 +0000)] 
Update the port knocking SPA document to have more details. Still needs a packet filter.

svn:r19356

17 years agoA small set of ideas that Nick and Roger suggested I write up regarding bridge detection.
Jacob Appelbaum [Tue, 21 Apr 2009 04:06:49 +0000 (04:06 +0000)] 
A small set of ideas that Nick and Roger suggested I write up regarding bridge detection.

svn:r19355

17 years agoNew upstream version
Peter Palfrader [Thu, 16 Apr 2009 18:39:03 +0000 (18:39 +0000)] 
New upstream version

svn:r19340

17 years agoDocumentation fix.
Karsten Loesing [Thu, 16 Apr 2009 08:25:36 +0000 (08:25 +0000)] 
Documentation fix.

svn:r19337

17 years agoUpdate to the "April 3 2009" ip-to-country file. tor-0.2.1.14-rc
Roger Dingledine [Sun, 12 Apr 2009 08:09:38 +0000 (08:09 +0000)] 
Update to the "April 3 2009" ip-to-country file.

svn:r19307

17 years agobump to 0.2.1.14-rc
Roger Dingledine [Sun, 12 Apr 2009 08:02:14 +0000 (08:02 +0000)] 
bump to 0.2.1.14-rc

svn:r19306

17 years agoRaise the minimum bandwidth to be a relay from 20000 bytes to 20480
Roger Dingledine [Sun, 12 Apr 2009 07:56:58 +0000 (07:56 +0000)] 
Raise the minimum bandwidth to be a relay from 20000 bytes to 20480
bytes (aka 20KB/s), to match our documentation. Also update
directory authorities so they always assign the Fast flag to relays
with 20KB/s of capacity. Now people running relays won't suddenly
find themselves not seeing any use, if the network gets faster
on average.

svn:r19305

17 years agoput HashedControlPassword and CookieAuthentication explicitly in the
Roger Dingledine [Sun, 12 Apr 2009 06:45:46 +0000 (06:45 +0000)] 
put HashedControlPassword and CookieAuthentication explicitly in the
torrc.sample.in, so unix people don't shoot themselves in the foot.

svn:r19304

17 years agoFor belt-and-suspenders, relays that don't set Address in their config
Roger Dingledine [Sat, 11 Apr 2009 19:40:42 +0000 (19:40 +0000)] 
For belt-and-suspenders, relays that don't set Address in their config
now avoid using begin_dir for all direct connections.

svn:r19296

17 years agomake it compile too
Roger Dingledine [Sat, 11 Apr 2009 19:22:38 +0000 (19:22 +0000)] 
make it compile too

svn:r19295

17 years agoplay make-believe that addr isn't a uint32_t
Roger Dingledine [Sat, 11 Apr 2009 18:21:18 +0000 (18:21 +0000)] 
play make-believe that addr isn't a uint32_t

svn:r19293

17 years agofix the same bug in two more locations (thanks to lark for prodding
Roger Dingledine [Sat, 11 Apr 2009 18:19:57 +0000 (18:19 +0000)] 
fix the same bug in two more locations (thanks to lark for prodding
me further)

svn:r19292

17 years agoFinally fix the bug where dynamic-IP relays disappear when their
Roger Dingledine [Sat, 11 Apr 2009 18:09:15 +0000 (18:09 +0000)] 
Finally fix the bug where dynamic-IP relays disappear when their
IP address changes: directory mirrors were mistakenly telling them
their old address if they asked via begin_dir, so they never got
an accurate answer about their new address, so they just vanished
after a day. Should fix bugs 827, 883, and 900 -- but alas, only
after every directory mirror has upgraded.

svn:r19291

17 years agoonly log that at loglevel notice if there's a problem with the
Roger Dingledine [Sat, 11 Apr 2009 12:06:27 +0000 (12:06 +0000)] 
only log that at loglevel notice if there's a problem with the
version. otherwise there's no reason to tell the user we're
doing behind-the-scenes cleaning.

svn:r19288

17 years agoClients replace entry guards that were chosen more than a few months
Roger Dingledine [Sat, 11 Apr 2009 12:00:18 +0000 (12:00 +0000)] 
Clients replace entry guards that were chosen more than a few months
ago. This change should significantly improve client performance,
especially once more people upgrade, since relays that have been
a guard for a long time are currently overloaded.

svn:r19287

17 years agogeneral changelog cleanups
Roger Dingledine [Sat, 11 Apr 2009 04:31:46 +0000 (04:31 +0000)] 
general changelog cleanups

svn:r19273

17 years agofix a comment
Roger Dingledine [Sat, 11 Apr 2009 00:40:11 +0000 (00:40 +0000)] 
fix a comment

svn:r19262

17 years agoa changelog entry for r19259
Roger Dingledine [Sat, 11 Apr 2009 00:24:06 +0000 (00:24 +0000)] 
a changelog entry for r19259

svn:r19261

17 years agoIf the bridge config line doesn't specify a port, assume 443.
Roger Dingledine [Sat, 11 Apr 2009 00:16:05 +0000 (00:16 +0000)] 
If the bridge config line doesn't specify a port, assume 443.
This makes bridge lines a bit smaller and easier for users to
understand.

Also, remove a duplicate changelog entry from the past.

svn:r19260

17 years agoDirectory authorities should accept a descriptor as changed when the relay operator...
Sebastian Hahn [Sat, 11 Apr 2009 00:12:52 +0000 (00:12 +0000)] 
Directory authorities should accept a descriptor as changed when the relay operator changed the bandwithrate or bandwithburst settings.

svn:r19259

17 years agocommit sebastian's patch to not require contactinfo when testingtornetwork
Roger Dingledine [Fri, 10 Apr 2009 07:02:19 +0000 (07:02 +0000)] 
commit sebastian's patch to not require contactinfo when testingtornetwork
is set

svn:r19257

17 years agoFix a few crash bugs related to malormed descriptors. Lark found one; fuzzing found...
Nick Mathewson [Thu, 9 Apr 2009 19:58:16 +0000 (19:58 +0000)] 
Fix a few crash bugs related to malormed descriptors.  Lark found one; fuzzing found the rest.

svn:r19250

17 years agoupdate to openssl 0.9.8k
Andrew Lewman [Thu, 9 Apr 2009 04:11:13 +0000 (04:11 +0000)] 
update to openssl 0.9.8k

svn:r19248

17 years agolog more verbosely when we accept or decline a router descriptor,
Roger Dingledine [Wed, 1 Apr 2009 13:02:04 +0000 (13:02 +0000)] 
log more verbosely when we accept or decline a router descriptor,
to help track whether we received them when a relay operator claims
they got sent.

svn:r19213

17 years agoif a directory authority declines our server descriptors because it's
Roger Dingledine [Tue, 31 Mar 2009 05:24:36 +0000 (05:24 +0000)] 
if a directory authority declines our server descriptors because it's
not new enough, write that in the logs. might be helpful one day.

svn:r19195

17 years agoLimit uploaded directory documents to be 16M rather than 500K.
Roger Dingledine [Tue, 31 Mar 2009 04:03:37 +0000 (04:03 +0000)] 
Limit uploaded directory documents to be 16M rather than 500K.
The directory authorities were refusing v3 consensus votes from
other authorities, since the votes are now 504K. Fixes bug 959;
bugfix on 0.0.2pre17 (where we raised it from 50K to 500K ;).

svn:r19194

17 years agoDirectory authorities should never send a 503 "busy" response to
Roger Dingledine [Tue, 31 Mar 2009 01:29:07 +0000 (01:29 +0000)] 
Directory authorities should never send a 503 "busy" response to
requests for votes or keys. Bugfix on 0.2.0.8-alpha; exposed by
bug 959.

svn:r19189

17 years agomove some of the items out of the 'phase one' section
Roger Dingledine [Sun, 29 Mar 2009 09:17:11 +0000 (09:17 +0000)] 
move some of the items out of the 'phase one' section

svn:r19180

17 years agoput in the performance todo items that i marked as high-priority in
Roger Dingledine [Sun, 29 Mar 2009 08:34:35 +0000 (08:34 +0000)] 
put in the performance todo items that i marked as high-priority in
the projects/performance/perf-todo file.

svn:r19178

17 years agonote some progress we've made on a few of them
Roger Dingledine [Sun, 29 Mar 2009 03:13:42 +0000 (03:13 +0000)] 
note some progress we've made on a few of them

svn:r19177

17 years agoput the 'phase 2' preliminary todo items in place
Roger Dingledine [Sun, 29 Mar 2009 03:11:19 +0000 (03:11 +0000)] 
put the 'phase 2' preliminary todo items in place

svn:r19176

17 years agoAdd proposal 159 to index.
Karsten Loesing [Mon, 23 Mar 2009 17:24:48 +0000 (17:24 +0000)] 
Add proposal 159 to index.

svn:r19114

17 years agotypo
Roger Dingledine [Mon, 23 Mar 2009 00:30:48 +0000 (00:30 +0000)] 
typo

svn:r19109

17 years agoActually do that memarea_strndup fix right. Not only must you not examine unmapped...
Nick Mathewson [Sat, 21 Mar 2009 16:01:52 +0000 (16:01 +0000)] 
Actually do that memarea_strndup fix right.  Not only must you not examine unmapped ram, but you also must not copy it.  From lark.

svn:r19095

17 years agofix from lark: make memarea_strndup() work even at the end of a mmap chunk. Bug...
Nick Mathewson [Sat, 21 Mar 2009 11:52:53 +0000 (11:52 +0000)] 
fix from lark: make memarea_strndup() work even at the end of a mmap chunk.  Bug was harmless for now, I think.

svn:r19094

17 years agofigure out version numbers for ostag in suse and redflag
Andrew Lewman [Wed, 18 Mar 2009 22:42:33 +0000 (22:42 +0000)] 
figure out version numbers for ostag in suse and redflag

svn:r19083

17 years agotag redflag linux rpms correctly.
Andrew Lewman [Wed, 18 Mar 2009 20:26:01 +0000 (20:26 +0000)] 
tag redflag linux rpms correctly.

svn:r19082

17 years agoWhitespace cleanup.
Nick Mathewson [Wed, 18 Mar 2009 19:30:37 +0000 (19:30 +0000)] 
Whitespace cleanup.

svn:r19080

17 years agoDo not generate the non-verbose circuit path when generating a circuit event unless...
Nick Mathewson [Wed, 18 Mar 2009 19:30:34 +0000 (19:30 +0000)] 
Do not generate the non-verbose circuit path when generating a circuit event unless we will use it.

svn:r19079

17 years agoIf we have a routerstatus but no routerinfo to name a router, use the routerstatus...
Nick Mathewson [Wed, 18 Mar 2009 19:30:30 +0000 (19:30 +0000)] 
If we have a routerstatus but no routerinfo to name a router, use the routerstatus instead when generating circuit events.  Also refactor a little.

svn:r19078

17 years agoAdd a function to get a LongName from a routerstatus. Needed for partial bug 941...
Nick Mathewson [Wed, 18 Mar 2009 19:30:26 +0000 (19:30 +0000)] 
Add a function to get a LongName from a routerstatus.  Needed for partial bug 941 fix.

svn:r19077

17 years agoClarify that the nickname part of a LongName is indeed optional.
Nick Mathewson [Wed, 18 Mar 2009 19:30:22 +0000 (19:30 +0000)] 
Clarify that the nickname part of a LongName is indeed optional.

svn:r19076

17 years agoAdd some asserts to try to catch bug 930
Nick Mathewson [Wed, 18 Mar 2009 15:12:56 +0000 (15:12 +0000)] 
Add some asserts to try to catch bug 930

svn:r19074

17 years agoDon't double-free successful_uploads.
Nick Mathewson [Wed, 18 Mar 2009 14:35:24 +0000 (14:35 +0000)] 
Don't double-free successful_uploads.

When we used smartlist_free to free the list of succesful uploads
because we had succeeded in uploading everywhere, we did not actually
set the successful_uploads field to NULL, so later it would get freed
again in rend_service_descriptor_free.  Fix for bug 948; bug
introduced in 0.2.1.6-alpha.

svn:r19073

17 years agoFree very-old descriptors that we do not want to add. Fix for bug 672. Backport...
Nick Mathewson [Mon, 16 Mar 2009 18:47:45 +0000 (18:47 +0000)] 
Free very-old descriptors that we do not want to add.  Fix for bug 672.  Backport candidate

svn:r19057

17 years agoThe override file is always correct
Peter Palfrader [Mon, 16 Mar 2009 17:08:45 +0000 (17:08 +0000)] 
The override file is always correct

svn:r19053

17 years agoUpdate translation doc with info on two options for new
Mike Perry [Mon, 16 Mar 2009 07:56:01 +0000 (07:56 +0000)] 
Update translation doc with info on two options for new
string creation workflow. It is unclear at this point
which of the two are the lesser evil, so both are included.

svn:r19037