]> git.ipfire.org Git - thirdparty/tor.git/log
thirdparty/tor.git
16 years agoDo the tor_assert hack everywhere in test_crypto_aes_iv.
Nick Mathewson [Tue, 13 Jan 2009 14:43:46 +0000 (14:43 +0000)] 
Do the tor_assert hack everywhere in test_crypto_aes_iv.

svn:r18095

16 years agoFix a leak memory on the failing case of test_memeq_hex
Nick Mathewson [Tue, 13 Jan 2009 14:43:43 +0000 (14:43 +0000)] 
Fix a leak memory on the failing case of test_memeq_hex

svn:r18094

16 years agoForward port 03_tor_manpage_in_section_8.dpatch
Peter Palfrader [Sun, 11 Jan 2009 11:12:04 +0000 (11:12 +0000)] 
Forward port 03_tor_manpage_in_section_8.dpatch

svn:r18074

16 years agoNew alpha release
Peter Palfrader [Sun, 11 Jan 2009 11:11:47 +0000 (11:11 +0000)] 
New alpha release

svn:r18073

16 years agoFix a harmless-to-us bug in ht.h.
Nick Mathewson [Sat, 10 Jan 2009 14:40:43 +0000 (14:40 +0000)] 
Fix a harmless-to-us bug in ht.h.
There was a field that _HT_FOI_INSERT was never setting. Everything that calls _HT_FOI_INSERT was setting it via tor_malloc_zero, but that's fragile.

svn:r18064

16 years agoFix ServerDNSRandomizeCase so that setting it to 0 works.
Nick Mathewson [Sat, 10 Jan 2009 14:38:14 +0000 (14:38 +0000)] 
Fix ServerDNSRandomizeCase so that setting it to 0 works.
Bugfix on 0.2.1.7-alpha.  Backport candidate. Fixes bug 905.

svn:r18063

16 years agoAUTHENTICATE is really mandatory. No authentication is not quite the default.
Nick Mathewson [Thu, 8 Jan 2009 14:07:05 +0000 (14:07 +0000)] 
AUTHENTICATE is really mandatory.  No authentication is not quite the default.

svn:r18024

16 years agoUpdate wix installer ids for 0.2.1.10-alpha MSI package.
Martin Peck [Thu, 8 Jan 2009 11:25:30 +0000 (11:25 +0000)] 
Update wix installer ids for 0.2.1.10-alpha MSI package.

svn:r18019

16 years agoMake the last bunch of libevent version detection silliness work on macosx.
Nick Mathewson [Wed, 7 Jan 2009 21:13:02 +0000 (21:13 +0000)] 
Make the last bunch of libevent version detection silliness work on macosx.

svn:r18015

16 years agoCheck that Libevent header version matches Libevent library version.
Nick Mathewson [Wed, 7 Jan 2009 21:05:02 +0000 (21:05 +0000)] 
Check that Libevent header version matches Libevent library version.
Unfortunately, old Libevents don't _put_ a version in their headers, so
this can get a little tricky.  Fortunately, the only binary-compatibility
issue we care about is the size of struct event.  Even more fortunately,
Libevent 2.0 will let us keep binary compatiblity forever by letting us
decouple ourselves from the structs, if we like.

svn:r18014

16 years agoDiscard router descriptors as we load them if they are more than
Roger Dingledine [Wed, 7 Jan 2009 02:13:42 +0000 (02:13 +0000)] 
Discard router descriptors as we load them if they are more than
five days old. Otherwise if Tor is off for a long time and then
starts with cached descriptors, it will try to use the onion
keys in those obsolete descriptors when building circuits. Bugfix
on 0.2.0.x. Fixes bug 887.

svn:r17993

16 years agoactually find the right identity_digest to blame when we get
Roger Dingledine [Wed, 7 Jan 2009 00:30:12 +0000 (00:30 +0000)] 
actually find the right identity_digest to blame when we get
a destroy cell from our first hop.

svn:r17991

16 years agowell, that was a disaster. this time for sure.
Roger Dingledine [Tue, 6 Jan 2009 22:20:31 +0000 (22:20 +0000)] 
well, that was a disaster. this time for sure.

svn:r17989

16 years agono actual changes besides a seg fault on startup, i meant
Roger Dingledine [Tue, 6 Jan 2009 21:42:05 +0000 (21:42 +0000)] 
no actual changes besides a seg fault on startup, i meant

svn:r17986

16 years agosimplify. no actual changes i think. ;)
Roger Dingledine [Tue, 6 Jan 2009 21:31:46 +0000 (21:31 +0000)] 
simplify. no actual changes i think. ;)

svn:r17985

16 years agoWhen our circuit fails at the first hop (e.g. we get a destroy
Roger Dingledine [Tue, 6 Jan 2009 21:28:48 +0000 (21:28 +0000)] 
When our circuit fails at the first hop (e.g. we get a destroy
cell back), avoid using that OR connection anymore, and also
tell all the one-hop directory requests waiting for it that they
should fail. Bugfix on 0.2.1.3-alpha.

svn:r17984

16 years agoMake outgoing DNS requests respect OutboundBindAddress.
Nick Mathewson [Tue, 6 Jan 2009 20:50:55 +0000 (20:50 +0000)] 
Make outgoing DNS requests respect OutboundBindAddress.
Fixes the bug part of bug 789.

svn:r17983

16 years agoNote a problem in the interface tor_addr_to_sockaddr.
Nick Mathewson [Tue, 6 Jan 2009 20:50:51 +0000 (20:50 +0000)] 
Note a problem in the interface tor_addr_to_sockaddr.

svn:r17982

16 years agook, all set to destabilize it again
Roger Dingledine [Tue, 6 Jan 2009 20:38:05 +0000 (20:38 +0000)] 
ok, all set to destabilize it again

svn:r17980

16 years agobump to 0.2.1.10-alpha, reorganize the changelog tor-0.2.1.10-alpha
Roger Dingledine [Tue, 6 Jan 2009 20:11:47 +0000 (20:11 +0000)] 
bump to 0.2.1.10-alpha, reorganize the changelog

svn:r17974

16 years agoclarify which relay gave us the good netinfo cell
Roger Dingledine [Tue, 6 Jan 2009 19:21:47 +0000 (19:21 +0000)] 
clarify which relay gave us the good netinfo cell

svn:r17973

16 years agoand mention the bug # in the changelog
Roger Dingledine [Tue, 6 Jan 2009 19:07:21 +0000 (19:07 +0000)] 
and mention the bug # in the changelog

svn:r17971

16 years agoWhen a relay gets a create cell it can't decrypt (e.g. because it's
Roger Dingledine [Tue, 6 Jan 2009 19:03:15 +0000 (19:03 +0000)] 
When a relay gets a create cell it can't decrypt (e.g. because it's
using the wrong onion key), we were dropping it and letting the
client time out. Now actually answer with a destroy cell. Bugfix
on 0.0.2pre8.

svn:r17970

16 years agoJan 06 13:03:57.309 [info] networkstatus_set_current_consensus(): Loaded an obsolete...
Roger Dingledine [Tue, 6 Jan 2009 18:06:49 +0000 (18:06 +0000)] 
Jan 06 13:03:57.309 [info] networkstatus_set_current_consensus(): Loaded an obsolete consensus. Discarding.
Jan 06 13:03:57.309 [warn] Couldn't load consensus networkstatus from "bridge/cached-consensus"

doesn't need a real warn.

svn:r17969

16 years agoUpdate to the "December 19 2008" ip-to-country file
Roger Dingledine [Tue, 6 Jan 2009 18:03:01 +0000 (18:03 +0000)] 
Update to the "December 19 2008" ip-to-country file

svn:r17968

16 years agoClean up (and mark for 0.2.2.) comments relating to non-beauty of current bug-743...
Nick Mathewson [Tue, 6 Jan 2009 17:37:22 +0000 (17:37 +0000)] 
Clean up (and mark for 0.2.2.) comments relating to non-beauty of current bug-743 fix.

svn:r17966

16 years agoSlightly better messages on ControlListenAddress 0.0.0.0
Nick Mathewson [Tue, 6 Jan 2009 16:57:42 +0000 (16:57 +0000)] 
Slightly better messages on ControlListenAddress 0.0.0.0

svn:r17963

16 years agospecify it was a bugfix on something from 2007
Roger Dingledine [Tue, 6 Jan 2009 16:11:10 +0000 (16:11 +0000)] 
specify it was a bugfix on something from 2007

svn:r17960

16 years agoPrevent bridge relays from serving their 'extrainfo' document
Roger Dingledine [Tue, 6 Jan 2009 16:03:38 +0000 (16:03 +0000)] 
Prevent bridge relays from serving their 'extrainfo' document
to anybody who asks, now that extrainfo docs include potentially
sensitive aggregated client geoip summaries.

svn:r17958

16 years agoMake dirserv_get_routerdesc_fingerprints() treat extrainfos with send_unencrypted...
Nick Mathewson [Tue, 6 Jan 2009 15:37:51 +0000 (15:37 +0000)] 
Make dirserv_get_routerdesc_fingerprints() treat extrainfos with send_unencrypted==0 correctly.  Irrelevant, since we will soon never send them at all.

svn:r17953

16 years agoBugfix on r13098. Backport candidate.
Roger Dingledine [Tue, 6 Jan 2009 07:13:24 +0000 (07:13 +0000)] 
Bugfix on r13098. Backport candidate.
When we made bridge authorities stop serving bridge descriptors over
unencrypted links, we also broke DirPort reachability testing for
bridges. So bridges with a non-zero DirPort were printing spurious
warns to their logs. Bugfix on 0.2.0.16-alpha. Fixes bug 709.

svn:r17945

16 years agothe new alpha release is coming out this week sometime. too many
Roger Dingledine [Tue, 6 Jan 2009 03:57:06 +0000 (03:57 +0000)] 
the new alpha release is coming out this week sometime. too many
bugs fixed to keep waiting. i still need to track down bug 709
first though, since i am increasingly thinking it's a real bug.

svn:r17944

16 years agoapply a variant of rovv's bug 872 patch, and document that we want a better solution...
Nick Mathewson [Mon, 5 Jan 2009 20:52:14 +0000 (20:52 +0000)] 
apply a variant of rovv's bug 872 patch, and document that we want a better solution for 0.2.2.x.

svn:r17924

16 years agoBridge relays that had DirPort set to 0 would stop fetching
Roger Dingledine [Mon, 5 Jan 2009 18:54:11 +0000 (18:54 +0000)] 
Bridge relays that had DirPort set to 0 would stop fetching
descriptors shortly after startup, and then briefly resume
after a new bandwidth test and/or after publishing a new bridge
descriptor. Bridge users that try to bootstrap from them would
get a recent networkstatus but would get descriptors from up to
18 hours earlier, meaning most of the descriptors were obsolete
already. Reported by Tas; bugfix on 0.2.0.13-alpha.

svn:r17920

16 years agoIf the cached networkstatus consensus is more than five days old,
Roger Dingledine [Mon, 5 Jan 2009 16:56:11 +0000 (16:56 +0000)] 
If the cached networkstatus consensus is more than five days old,
discard it rather than trying to use it. In theory it could
be useful because it lists alternate directory mirrors, but in
practice it just means we spend many minutes trying directory
mirrors that are long gone from the network. Helps bug 887 a bit;
bugfix on 0.2.0.x.

svn:r17917

16 years agomark what version of tor mike's bug was in
Roger Dingledine [Mon, 5 Jan 2009 14:32:58 +0000 (14:32 +0000)] 
mark what version of tor mike's bug was in

svn:r17912

16 years agoPreserve reporting of stream end reasons to the local control
Mike Perry [Mon, 5 Jan 2009 14:14:57 +0000 (14:14 +0000)] 
Preserve reporting of stream end reasons to the local control
port.  They were lost in the changes for Proposal 148.

svn:r17911

16 years agoBetter explain the PublishServerDescriptor option in the manpage
Sebastian Hahn [Mon, 5 Jan 2009 13:50:02 +0000 (13:50 +0000)] 
Better explain the PublishServerDescriptor option in the manpage

PublishServerDescriptor 1 means "publish to the relevant authority", not "publish to v2,v3 authorities". Also adjust the time in the manpage to January 2009

svn:r17909

16 years agoUse doxygen-style comments instead of C-style comments.
Karsten Loesing [Sun, 4 Jan 2009 23:44:43 +0000 (23:44 +0000)] 
Use doxygen-style comments instead of C-style comments.

svn:r17893

16 years agoAnother fun openbsd warning fix. On ioerror's computer at least, they redefined...
Nick Mathewson [Sun, 4 Jan 2009 23:15:42 +0000 (23:15 +0000)] 
Another fun openbsd warning fix.  On ioerror's computer at least, they redefined an unsigned field in zlib.h to be signed.  I am quite sure this makes me more secure somehow.

svn:r17892

16 years agoOpenBSD malloc.h believes that you should be able to detect headers with autoconf...
Nick Mathewson [Sun, 4 Jan 2009 22:47:42 +0000 (22:47 +0000)] 
OpenBSD malloc.h believes that you should be able to detect headers with autoconf, or build without warnings, but not both.  So never include malloc.h on OpenBSD.  Backport candidate.

svn:r17891

16 years agoDocument some variables used for generating statistics on usage of V0 hidden service...
Karsten Loesing [Sun, 4 Jan 2009 19:59:35 +0000 (19:59 +0000)] 
Document some variables used for generating statistics on usage of V0 hidden service directory authorities.

svn:r17890

16 years agoAnother round of downgrading removing or postponing XXXX021 issues. Some remain...
Nick Mathewson [Sun, 4 Jan 2009 19:47:21 +0000 (19:47 +0000)] 
Another round of downgrading removing or postponing XXXX021 issues.  Some remain, though.

svn:r17888

16 years agoFix the oldest bug in a while: stop accepting 1.2.3 as a valid IPv4 address on any...
Nick Mathewson [Sun, 4 Jan 2009 19:47:17 +0000 (19:47 +0000)] 
Fix the oldest bug in a while: stop accepting 1.2.3 as a valid IPv4 address on any platform.

svn:r17887

16 years agoFix an xxx021: do not remove routerinfos as too old and unlisted unless we have a...
Nick Mathewson [Sun, 4 Jan 2009 19:47:12 +0000 (19:47 +0000)] 
Fix an xxx021: do not remove routerinfos as too old and unlisted unless we have a consensus.  Backport candidate.

svn:r17886

16 years agoCommit changelog from last year (0.2.1.9-alpha-1)
Peter Palfrader [Sun, 4 Jan 2009 19:30:04 +0000 (19:30 +0000)] 
Commit changelog from last year (0.2.1.9-alpha-1)

svn:r17884

16 years agoRemove svn $Id$s from our source, and remove tor --version --version.
Nick Mathewson [Sun, 4 Jan 2009 00:35:51 +0000 (00:35 +0000)] 
Remove svn $Id$s from our source, and remove tor --version --version.
The subversion $Id$ fields made every commit force a rebuild of
whatever file got committed.  They were not actually useful for
telling the version of Tor files in the wild.

svn:r17867

16 years agoList places where the hidden service protocol uses SHA-1.
Karsten Loesing [Sun, 4 Jan 2009 00:21:28 +0000 (00:21 +0000)] 
List places where the hidden service protocol uses SHA-1.

svn:r17866

16 years agoAnother attempt at fixing coverity scan CID 361.
Nick Mathewson [Sat, 3 Jan 2009 23:06:12 +0000 (23:06 +0000)] 
Another attempt at fixing coverity scan CID 361.

svn:r17863

16 years agoMention that our documentation is better in 0.2.1.10-x.
Nick Mathewson [Sat, 3 Jan 2009 23:06:08 +0000 (23:06 +0000)] 
Mention that our documentation is better in 0.2.1.10-x.

svn:r17862

16 years agoOnly set sin_len/sin6_len when they exist.
Nick Mathewson [Fri, 2 Jan 2009 20:57:10 +0000 (20:57 +0000)] 
Only set sin_len/sin6_len when they exist.

svn:r17851

16 years agotake out my IMPOSSIBLE_TO_DOWNLOAD+1 hack
Roger Dingledine [Fri, 2 Jan 2009 20:46:32 +0000 (20:46 +0000)] 
take out my IMPOSSIBLE_TO_DOWNLOAD+1 hack

svn:r17850

16 years agoTry harder to make sure we zero-out the extraneous sockaddr fields and that we set...
Nick Mathewson [Fri, 2 Jan 2009 20:39:38 +0000 (20:39 +0000)] 
Try harder to make sure we zero-out the extraneous sockaddr fields and that we set sockaddr_len.  Conceivably a backport candidate, though nothing has yet been sen to break.

svn:r17849

16 years agoNote another place we use sha1
Nick Mathewson [Wed, 31 Dec 2008 18:25:30 +0000 (18:25 +0000)] 
Note another place we use sha1

svn:r17836

16 years agoIncrement n_download_failures up to IMPOSSIBLE_TO_DOWNLOAD-1, not up to IMPOSSIBLE_TO...
Nick Mathewson [Wed, 31 Dec 2008 17:22:18 +0000 (17:22 +0000)] 
Increment n_download_failures up to IMPOSSIBLE_TO_DOWNLOAD-1, not up to IMPOSSIBLE_TO_DOWNLOAD.

svn:r17835

16 years agoswitch over the bridge descriptor download mechanism to
Roger Dingledine [Wed, 31 Dec 2008 14:19:57 +0000 (14:19 +0000)] 
switch over the bridge descriptor download mechanism to
use the same download mechanism as other places.

i had to make an ugly hack around "IMPOSSIBLE_TO_DOWNLOAD+1".
we should unhack that sometime.

svn:r17834

16 years agofix a bug in download failure logging; don't do this "stop
Roger Dingledine [Wed, 31 Dec 2008 14:11:46 +0000 (14:11 +0000)] 
fix a bug in download failure logging; don't do this "stop
downloading forever" thing with bridge descriptor downloads.

svn:r17833

16 years agoNew proposal "idea" listing all the places I could think of that we use SHA-1.
Nick Mathewson [Tue, 30 Dec 2008 17:15:27 +0000 (17:15 +0000)] 
New proposal "idea" listing all the places I could think of that we use SHA-1.

svn:r17831

16 years agoDebug code for task 878.
Karsten Loesing [Tue, 30 Dec 2008 11:43:50 +0000 (11:43 +0000)] 
Debug code for task 878.

svn:r17828

16 years agoRemove some code that is #ifdefed out, and that we no longer seem to use, if we ever...
Nick Mathewson [Tue, 30 Dec 2008 04:16:49 +0000 (04:16 +0000)] 
Remove some code that is #ifdefed out, and that we no longer seem to use, if we ever did.

svn:r17827

16 years agoIndeed, arma was right. There was one ntohl too many in debug_ntop in eventdns.c
Nick Mathewson [Tue, 30 Dec 2008 04:08:45 +0000 (04:08 +0000)] 
Indeed, arma was right.  There was one ntohl too many in debug_ntop in eventdns.c

svn:r17826

16 years agopoint out the bug location to nick. unless i'm wrong.
Roger Dingledine [Tue, 30 Dec 2008 02:36:11 +0000 (02:36 +0000)] 
point out the bug location to nick. unless i'm wrong.

svn:r17825

16 years agofor some reason nickm removed this major feature from
Roger Dingledine [Tue, 30 Dec 2008 02:03:23 +0000 (02:03 +0000)] 
for some reason nickm removed this major feature from
the 0.2.1.3-alpha changelog. put it back.

svn:r17824

16 years agoDowngrade some xxx021s, comment more on others, etc
Nick Mathewson [Mon, 29 Dec 2008 20:17:24 +0000 (20:17 +0000)] 
Downgrade some xxx021s, comment more on others, etc

svn:r17823

16 years agoActually log reasons on unusable guards again.
Nick Mathewson [Mon, 29 Dec 2008 20:17:20 +0000 (20:17 +0000)] 
Actually log reasons on unusable guards again.

svn:r17822

16 years agoSpec compliance: if weighted MTBF of a router is at least 5 days, always vote it...
Nick Mathewson [Mon, 29 Dec 2008 20:17:18 +0000 (20:17 +0000)] 
Spec compliance: if weighted MTBF of a router is at least 5 days, always vote it as Stable.

svn:r17821

16 years agoRefactor sockaddr family match check into a new function
Nick Mathewson [Mon, 29 Dec 2008 19:57:17 +0000 (19:57 +0000)] 
Refactor sockaddr family match check into a new function

svn:r17820

16 years agoExplain why the XXXX021 in connection_or.c was there, and explain why it doesn' actua...
Nick Mathewson [Mon, 29 Dec 2008 19:57:13 +0000 (19:57 +0000)] 
Explain why the XXXX021 in connection_or.c was there, and explain why it doesn' actually need to get fixed in 0.2.1, I think.

svn:r17819

16 years agoReplace kludgy assert with something a little less awful, and avoid needless strlcpy...
Nick Mathewson [Mon, 29 Dec 2008 19:57:08 +0000 (19:57 +0000)] 
Replace kludgy assert with something a little less awful, and avoid needless strlcpy in rend descriptor encoding.

svn:r17818

16 years agoDowngrade the last xxx021 in routerparse. The duplicate code stands for now.
Nick Mathewson [Mon, 29 Dec 2008 19:57:04 +0000 (19:57 +0000)] 
Downgrade the last xxx021 in routerparse.  The duplicate code stands for now.

svn:r17817

16 years agoDocument RELAY_EARLY cells in tor-spec.txt: proposal 110 is mostly implemented, and...
Nick Mathewson [Mon, 29 Dec 2008 19:55:23 +0000 (19:55 +0000)] 
Document RELAY_EARLY cells in tor-spec.txt: proposal 110 is mostly implemented, and the rest of it will need to wait until all current Tor 0.2.0.x clients are obsolete to close.

svn:r17816

16 years agoDo not not cannibalize a circuit that has run out of RELAY_EARLY cells. Partial...
Nick Mathewson [Mon, 29 Dec 2008 19:55:17 +0000 (19:55 +0000)] 
Do not not cannibalize a circuit that has run out of RELAY_EARLY cells.  Partial bug 878 fix.

svn:r17815

16 years agoDo not cannibalize a circuit that has run out of RELAY_EARLY cells. Partial bug...
Nick Mathewson [Mon, 29 Dec 2008 19:55:13 +0000 (19:55 +0000)] 
Do not cannibalize a circuit that has run out of RELAY_EARLY cells.  Partial bug 878 fix.

svn:r17814

16 years agoExtact parse-the-token-arguments to its own function, and change it to a single-pass...
Nick Mathewson [Mon, 29 Dec 2008 16:54:56 +0000 (16:54 +0000)] 
Extact parse-the-token-arguments to its own function, and change it to a single-pass algorithm.  This simplifies the parsing code and speeds it up a little.

svn:r17812

16 years agoRemove a call to find_whitespace_eos that didn't actually do anything.
Nick Mathewson [Mon, 29 Dec 2008 16:54:51 +0000 (16:54 +0000)] 
Remove a call to find_whitespace_eos that didn't actually do anything.

svn:r17811

16 years agoWhile I'm at it, refactor control.c a little so that the dead code no longer exists.
Nick Mathewson [Mon, 29 Dec 2008 14:21:25 +0000 (14:21 +0000)] 
While I'm at it, refactor control.c a little so that the dead code no longer exists.

svn:r17810

16 years agocoverity thinks it's dumb to check networkstatus_v2_list in one place and not another...
Nick Mathewson [Mon, 29 Dec 2008 14:06:47 +0000 (14:06 +0000)] 
coverity thinks it's dumb to check networkstatus_v2_list in one place and not another.  Coverity has a point, even though the check may be redundant. CID 361.

svn:r17809

16 years agoMore compilation tweaks on Android: fix two clear errors in our code that apparently...
Nick Mathewson [Mon, 29 Dec 2008 04:01:58 +0000 (04:01 +0000)] 
More compilation tweaks on Android: fix two clear errors in our code that apparently the compiler cares about there.

svn:r17808

16 years agoUpdate openssl version, suggest latest libevent.
Andrew Lewman [Mon, 29 Dec 2008 03:46:50 +0000 (03:46 +0000)] 
Update openssl version, suggest latest libevent.

svn:r17806

16 years agoUse a consistent naming standard for header file guard macros, taking care not to...
Nick Mathewson [Mon, 29 Dec 2008 02:21:02 +0000 (02:21 +0000)] 
Use a consistent naming standard for header file guard macros, taking care not to collide with any system headers.  This tripped us up on Android.

svn:r17805

16 years agoUse RSA_generate_key_ex where available.
Nick Mathewson [Mon, 29 Dec 2008 02:20:57 +0000 (02:20 +0000)] 
Use RSA_generate_key_ex where available.

svn:r17804

16 years agoSwitch address comparisons in policies to be exact rather than semantic. Until we...
Nick Mathewson [Mon, 29 Dec 2008 01:47:33 +0000 (01:47 +0000)] 
Switch address comparisons in policies to be exact rather than semantic.  Until we do ipv6 exit policies and until we know whether we even allow ::ffff:0:0/96 addresses, there is no point in doing "semantic" comparisons.  This was also showing up on oprofile.

svn:r17803

16 years agoRefactor some exit-policy-related functions that showed up in oprofile.
Nick Mathewson [Mon, 29 Dec 2008 01:47:28 +0000 (01:47 +0000)] 
Refactor some exit-policy-related functions that showed up in oprofile.
Specifically, split compare_tor_addr_to_addr_policy() from a loop with a bunch
of complicated ifs inside into some ifs, each with a simple loop.  Rearrange
router_find_exact_exit_enclave() to run a little faster.  Bizarrely,
router_policy_rejects_all() shows up on oprofile, so precalculate it per
routerinfo.

svn:r17802

16 years agoFix bug in recent address.c patch: actually set the value of address * to 0.0.0.0...
Nick Mathewson [Mon, 29 Dec 2008 01:30:35 +0000 (01:30 +0000)] 
Fix bug in recent address.c patch: actually set the value of address * to 0.0.0.0 as we did before.  This makes CMP_EXACT comparisons with bitmask 0 work on address * again.

svn:r17801

16 years agoFix compilation under gethostbyname-based systems.
Nick Mathewson [Sat, 27 Dec 2008 15:46:16 +0000 (15:46 +0000)] 
Fix compilation under gethostbyname-based systems.

svn:r17800

16 years agoremove the responsibility for setting listensocklen to the function that made the...
Nick Mathewson [Sat, 27 Dec 2008 15:46:13 +0000 (15:46 +0000)] 
remove the responsibility for setting listensocklen to the function that made the sockaddr

svn:r17799

16 years agopartial move to letting bridge descriptor fetches use our new (well,
Roger Dingledine [Sat, 27 Dec 2008 07:30:47 +0000 (07:30 +0000)] 
partial move to letting bridge descriptor fetches use our new (well,
new from their perspective) directory download schedule abstraction.

not done yet, but i'd better get this out of my sandbox before nick
does another sweeping change. :)

svn:r17798

16 years agominor todo updates
Roger Dingledine [Sat, 27 Dec 2008 07:28:47 +0000 (07:28 +0000)] 
minor todo updates

svn:r17797

16 years agoNew controller event "clients_seen" to report a geoip-based summary
Roger Dingledine [Sat, 27 Dec 2008 06:50:07 +0000 (06:50 +0000)] 
New controller event "clients_seen" to report a geoip-based summary
of which countries we've seen clients from recently. Now controllers
like Vidalia can show bridge operators that they're actually making
a difference.

svn:r17796

16 years agoRemove a deprecated controller alias; make another one generate a warning.
Nick Mathewson [Sat, 27 Dec 2008 00:20:08 +0000 (00:20 +0000)] 
Remove a deprecated controller alias; make another one generate a warning.

svn:r17793

16 years agoMake connection_create_listener take a sockaddr_len, not just a sockaddr whose length...
Nick Mathewson [Sat, 27 Dec 2008 00:20:06 +0000 (00:20 +0000)] 
Make connection_create_listener take a sockaddr_len, not just a sockaddr whose length we try to divine.

svn:r17792

16 years agoAdd some early checks to keep ipv6 addresses from failing by accident. Now, they...
Nick Mathewson [Fri, 26 Dec 2008 22:51:25 +0000 (22:51 +0000)] 
Add some early checks to keep ipv6 addresses from failing by accident.  Now, they fail on purpose, at least till 0.2.2.something.

svn:r17791

16 years agoRefactor tor_addr_t manipulation functions so that as few as possible look at the...
Nick Mathewson [Fri, 26 Dec 2008 21:26:05 +0000 (21:26 +0000)] 
Refactor tor_addr_t manipulation functions so that as few as possible look at the tor_addr_t representation.

svn:r17790

16 years agoget_interface_addr6(), and by extension get_interface_addr(), were pretty borked...
Nick Mathewson [Fri, 26 Dec 2008 21:26:03 +0000 (21:26 +0000)] 
get_interface_addr6(), and by extension get_interface_addr(), were pretty borked.  Copying a tor_addr_t from a sockaddr_storage using memcpy is a poor notion.

svn:r17789

16 years agoRefactor tor_addr_compare_masked() so that CMP_SEMANTIC makes more sense, and has...
Nick Mathewson [Fri, 26 Dec 2008 20:37:18 +0000 (20:37 +0000)] 
Refactor tor_addr_compare_masked() so that CMP_SEMANTIC makes more sense, and has decent semantics for maskbits; and so that CMP_EXACT works right for bits==0.

svn:r17788

16 years agoBump version to 0.2.1.9-alpha-dev
Nick Mathewson [Fri, 26 Dec 2008 17:39:22 +0000 (17:39 +0000)] 
Bump version to 0.2.1.9-alpha-dev

svn:r17787

16 years agoDocument and make more conservative the logic for picking the size of the bloom filte...
Nick Mathewson [Fri, 26 Dec 2008 17:35:36 +0000 (17:35 +0000)] 
Document and make more conservative the logic for picking the size of the bloom filter to use for retaining descriptors.

svn:r17786

16 years agoDocument our Bloom filter parameter choices.
Nick Mathewson [Fri, 26 Dec 2008 17:35:18 +0000 (17:35 +0000)] 
Document our Bloom filter parameter choices.

svn:r17785

16 years agoDocument the pk operation count fields in rephist.c, and move them into a struct...
Nick Mathewson [Fri, 26 Dec 2008 17:35:12 +0000 (17:35 +0000)] 
Document the pk operation count fields in rephist.c, and move them into a struct for clarity, and change their type to what we actually want to cast them to.

svn:r17784

16 years agoAdd more missing documentation, and correct an error in container.c documentation...
Nick Mathewson [Fri, 26 Dec 2008 17:35:08 +0000 (17:35 +0000)] 
Add more missing documentation, and correct an error in container.c documentation: Don't introduce two parameters called n when you're calling an algorithm O(n).

svn:r17783