]> git.ipfire.org Git - thirdparty/tor.git/log
thirdparty/tor.git
8 years agoMake some assertions nonfatal to help prevent bug23690 recurrence.
Nick Mathewson [Fri, 29 Sep 2017 14:05:35 +0000 (10:05 -0400)] 
Make some assertions nonfatal to help prevent bug23690 recurrence.

8 years agoMerge remote-tracking branch 'catalyst-oniongit/bug23691'
Nick Mathewson [Thu, 28 Sep 2017 23:07:23 +0000 (19:07 -0400)] 
Merge remote-tracking branch 'catalyst-oniongit/bug23691'

8 years agoMerge remote-tracking branch 'ahf-gitlab/bugs/23551'
Nick Mathewson [Thu, 28 Sep 2017 22:54:40 +0000 (18:54 -0400)] 
Merge remote-tracking branch 'ahf-gitlab/bugs/23551'

8 years agoMerge branch 'maint-0.3.1'
Nick Mathewson [Thu, 28 Sep 2017 22:54:18 +0000 (18:54 -0400)] 
Merge branch 'maint-0.3.1'

8 years agoFix leak in test_util.c
Taylor Yu [Thu, 28 Sep 2017 22:53:38 +0000 (17:53 -0500)] 
Fix leak in test_util.c

AddressSanitizer found a leak in test_util_decompress_dos_impl().
Fixes bug #23691.

8 years agoMerge remote-tracking branch 'ahf-gitlab/bugs/23551-maint-0.3.1' into maint-0.3.1
Nick Mathewson [Thu, 28 Sep 2017 22:53:37 +0000 (18:53 -0400)] 
Merge remote-tracking branch 'ahf-gitlab/bugs/23551-maint-0.3.1' into maint-0.3.1

8 years agoBetter error handling when trying to compress/decompress into empty buffer.
Alexander Færøy [Thu, 28 Sep 2017 14:46:10 +0000 (16:46 +0200)] 
Better error handling when trying to compress/decompress into empty buffer.

This patch ensures that we return TOR_COMPRESS_BUFFER_FULL in case we
have a input bytes left to process, but are out of output buffer or in
case we need to finish where the compression implementation might need
to write an epilogue.

See: https://bugs.torproject.org/23551

8 years agoEnable disabled test that was disabled due to bug #23551
Alexander Færøy [Thu, 28 Sep 2017 16:46:18 +0000 (18:46 +0200)] 
Enable disabled test that was disabled due to bug #23551

See: https://bugs.torproject.org/23551

8 years agoBetter error handling when trying to compress/decompress into empty buffer.
Alexander Færøy [Thu, 28 Sep 2017 14:46:10 +0000 (16:46 +0200)] 
Better error handling when trying to compress/decompress into empty buffer.

This patch ensures that we return TOR_COMPRESS_BUFFER_FULL in case we
have a input bytes left to process, but are out of output buffer or in
case we need to finish where the compression implementation might need
to write an epilogue.

See: https://bugs.torproject.org/23551

8 years agoFix whitespace issue in compress.c
Alexander Færøy [Thu, 28 Sep 2017 13:15:37 +0000 (15:15 +0200)] 
Fix whitespace issue in compress.c

8 years agoFix typo in buffers.c.
Alexander Færøy [Thu, 28 Sep 2017 13:14:50 +0000 (15:14 +0200)] 
Fix typo in buffers.c.

8 years agoImprove unit test coverage for compression code.
Nick Mathewson [Thu, 28 Sep 2017 16:20:02 +0000 (12:20 -0400)] 
Improve unit test coverage for compression code.

These tests try uncompressing garbage, verify that we won't
make compression bombs, and verify that we won't uncompress
compression bombs.

8 years agoFix spelling: compressing, not compresing
Nick Mathewson [Thu, 28 Sep 2017 16:17:34 +0000 (12:17 -0400)] 
Fix spelling: compressing, not compresing

8 years agoUnit test for case where %included dir exists but is unreadable
Nick Mathewson [Thu, 28 Sep 2017 13:49:30 +0000 (09:49 -0400)] 
Unit test for case where %included dir exists but is unreadable

8 years agounit test for config_lines_dup_and_filter
Nick Mathewson [Thu, 28 Sep 2017 13:40:17 +0000 (09:40 -0400)] 
unit test for config_lines_dup_and_filter

8 years agoCorrect docs for config_lines_dup_and_filter
Nick Mathewson [Thu, 28 Sep 2017 13:40:04 +0000 (09:40 -0400)] 
Correct docs for config_lines_dup_and_filter

8 years agoMove around some LCOV_EXCLs in src/common
Nick Mathewson [Thu, 28 Sep 2017 13:25:17 +0000 (09:25 -0400)] 
Move around some LCOV_EXCLs in src/common

Apparently, my compiler now generates coverage markers for
label-only lines, so we need to exclude those too if they are meant
to be unreachable.

8 years agoImprove cov-diff script to account for new files.
Nick Mathewson [Thu, 28 Sep 2017 12:53:16 +0000 (08:53 -0400)] 
Improve cov-diff script to account for new files.

8 years agoAdd unit test for buf_get_line().
Nick Mathewson [Thu, 28 Sep 2017 12:35:30 +0000 (08:35 -0400)] 
Add unit test for buf_get_line().

8 years agoNote an unreachable (?) section in buffers.c
Nick Mathewson [Thu, 28 Sep 2017 12:35:24 +0000 (08:35 -0400)] 
Note an unreachable (?) section in buffers.c

8 years agoLet's get 100% coverage for proto_http while we're at it.
Nick Mathewson [Thu, 28 Sep 2017 01:45:55 +0000 (21:45 -0400)] 
Let's get 100% coverage for proto_http while we're at it.

There was some coverage before, but it was mostly accidental.

8 years agoTest more error cases of our socks code.
Nick Mathewson [Wed, 27 Sep 2017 23:37:36 +0000 (19:37 -0400)] 
Test more error cases of our socks code.

Coverage is now respectable. :)

8 years agoAdd unit tests for SOCKS functions that parse server responses
Nick Mathewson [Wed, 27 Sep 2017 22:55:14 +0000 (18:55 -0400)] 
Add unit tests for SOCKS functions that parse server responses

We use these when we're acting as a SOCKS client, but we'd never
actually written tests for them :/

8 years agosocks5 unit tests: add check for ipv6 address type.
Nick Mathewson [Wed, 27 Sep 2017 15:23:09 +0000 (11:23 -0400)] 
socks5 unit tests: add check for ipv6 address type.

8 years agoSocks tests for bad socks5 username/passwd auth.
Nick Mathewson [Wed, 27 Sep 2017 15:16:29 +0000 (11:16 -0400)] 
Socks tests for bad socks5 username/passwd auth.

8 years agoMark some tests in parse_socks.c as unreachable (BUG, LCOV)
Nick Mathewson [Wed, 27 Sep 2017 15:15:03 +0000 (11:15 -0400)] 
Mark some tests in parse_socks.c as unreachable (BUG, LCOV)

These tests aren't reachable, given their actual arguments.  I'm
going to mark them as BUG(), and as unreachable with LCOV.

8 years agotesting: test many possible truncated SOCKS commands
Nick Mathewson [Wed, 27 Sep 2017 14:54:56 +0000 (10:54 -0400)] 
testing: test many possible truncated SOCKS commands

Many of the 'truncated command' paths in fetch_from_buf_socks() were
not reached by the tests.  This new unit test tries to check them
exhaustively.

8 years agoMove ext_or_command tests to test_proto_misc.c
Nick Mathewson [Wed, 27 Sep 2017 13:11:15 +0000 (09:11 -0400)] 
Move ext_or_command tests to test_proto_misc.c

No other code changes.

8 years agoAdd unit tests for var_cell and control0 proto functions.
Nick Mathewson [Wed, 27 Sep 2017 13:09:00 +0000 (09:09 -0400)] 
Add unit tests for var_cell and control0 proto functions.

8 years agoEat redundant semicolons
Taylor Yu [Tue, 26 Sep 2017 18:41:20 +0000 (13:41 -0500)] 
Eat redundant semicolons

8 years agofix the changes file
Nick Mathewson [Tue, 26 Sep 2017 18:51:35 +0000 (14:51 -0400)] 
fix the changes file

8 years agohs-v3: Don't non fatal assert if we can't get the intro extend info
David Goulet [Tue, 26 Sep 2017 17:55:51 +0000 (13:55 -0400)] 
hs-v3: Don't non fatal assert if we can't get the intro extend info

Fixes #23159.

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agoMerge branch 'ticket23602_029'
Nick Mathewson [Tue, 26 Sep 2017 17:02:26 +0000 (13:02 -0400)] 
Merge branch 'ticket23602_029'

8 years agoMerge branch 'typecheck4'
Nick Mathewson [Tue, 26 Sep 2017 16:26:01 +0000 (12:26 -0400)] 
Merge branch 'typecheck4'

8 years agochanges file for my confparse typechecking fun
Nick Mathewson [Mon, 25 Sep 2017 15:55:51 +0000 (11:55 -0400)] 
changes file for my confparse typechecking fun

8 years agoType bug in shared_random_state -- make sure Version is int.
Nick Mathewson [Mon, 25 Sep 2017 15:52:57 +0000 (11:52 -0400)] 
Type bug in shared_random_state -- make sure Version is int.

The confparse field has type UINT, which corresponds to an int
type.  We had uint32_t.

This shouldn't cause trouble in practice, since int happens to
4-bytes wide on every platform where an authority is running.  It's
still wrong, though.

8 years agoMake the TransProxyType field non-const
Nick Mathewson [Tue, 26 Sep 2017 16:25:07 +0000 (12:25 -0400)] 
Make the TransProxyType field non-const

The correct type for a STRING confparse value is char *, not const
char *.

8 years agoCorrect two state-file variable types.
Nick Mathewson [Mon, 25 Sep 2017 15:51:15 +0000 (11:51 -0400)] 
Correct two state-file variable types.

These should have been int, but we had listed them as unsigned.
That's an easy mistake to make, since "int" corresponds with either
INT or UINT in the configuration file.

This bug cannot have actually caused a problem in practice, since we
check those fields' values on load, and ensure that they are in
range 0..INT32_MAX.

8 years agoAdd test to make sure all confparse variables are well-typed
Nick Mathewson [Mon, 25 Sep 2017 15:08:11 +0000 (11:08 -0400)] 
Add test to make sure all confparse variables are well-typed

New approach, suggested by Taylor: During testing builds, we
initialize a union member of an appropriate pointer type with the
address of the member field we're trying to test, so we can make
sure that the compiler doesn't warn.

My earlier approach invoked undefined behavior.

8 years agoadd changes file for 23645
Nick Mathewson [Tue, 26 Sep 2017 12:47:02 +0000 (08:47 -0400)] 
add changes file for 23645

8 years agoMerge remote-tracking branch 'asn/ticket23645_032_01'
Nick Mathewson [Tue, 26 Sep 2017 12:55:59 +0000 (08:55 -0400)] 
Merge remote-tracking branch 'asn/ticket23645_032_01'

8 years agohs: Improve logging
David Goulet [Mon, 25 Sep 2017 20:31:43 +0000 (16:31 -0400)] 
hs: Improve logging

Also demote a log message that can occur under natural causes
(if the circuit subsystem is missing descriptors/consensus etc.).
The HS subsystem will naturally retry to connect to intro points,
so no need to make that log user-facing.

8 years agocircuit: Log n_circ_id and global identifier of HS circuits
David Goulet [Mon, 25 Sep 2017 20:20:15 +0000 (16:20 -0400)] 
circuit: Log n_circ_id and global identifier of HS circuits

So we can track them more easily in the logs and match any open/close/free
with those identifiers.

Part of #23645

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agocircuit: Log circuit identifiers when cannibalizing
David Goulet [Mon, 25 Sep 2017 20:08:11 +0000 (16:08 -0400)] 
circuit: Log circuit identifiers when cannibalizing

This removes the "nickname" of the cannibalized circuit last hop as it is
useless. It now logs the n_circ_id and global identifier so we can match it
with other logging statement.

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agocircuit: Keep a copy of n_circ_id for logging
David Goulet [Mon, 25 Sep 2017 20:04:30 +0000 (16:04 -0400)] 
circuit: Keep a copy of n_circ_id for logging

Prior to the log statement, the circuit n_circ_id value is zeroed so keep a
copy so we can log it at the end.

Part of #23645

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agocircuit: Closing reason are signed values
David Goulet [Mon, 25 Sep 2017 20:01:12 +0000 (16:01 -0400)] 
circuit: Closing reason are signed values

Part of #23645

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agoMerge branch 'bug22109'
Nick Mathewson [Mon, 25 Sep 2017 18:32:04 +0000 (14:32 -0400)] 
Merge branch 'bug22109'

8 years agoMake testing rend1_payload variables static
Nick Mathewson [Mon, 25 Sep 2017 17:26:39 +0000 (13:26 -0400)] 
Make testing rend1_payload variables static

This fixes a clang warning.

8 years agoMerge branch 'bug23539_032_01_squashed'
Nick Mathewson [Mon, 25 Sep 2017 15:12:56 +0000 (11:12 -0400)] 
Merge branch 'bug23539_032_01_squashed'

8 years agoWhitespace fix.
Nick Mathewson [Mon, 25 Sep 2017 15:12:45 +0000 (11:12 -0400)] 
Whitespace fix.

8 years agosched: make interval a plain int; initialize with macro
Matt Traudt [Mon, 25 Sep 2017 14:43:15 +0000 (10:43 -0400)] 
sched: make interval a plain int; initialize with macro

8 years agosched: Make KISTSchedRunInterval non negative
David Goulet [Fri, 22 Sep 2017 15:33:50 +0000 (11:33 -0400)] 
sched: Make KISTSchedRunInterval non negative

Fixes #23539.

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agofix changes file
Nick Mathewson [Mon, 25 Sep 2017 13:29:11 +0000 (09:29 -0400)] 
fix changes file

8 years agoMerge remote-tracking branch 'asn/ticket23420_032_01'
Nick Mathewson [Mon, 25 Sep 2017 13:24:36 +0000 (09:24 -0400)] 
Merge remote-tracking branch 'asn/ticket23420_032_01'

8 years agoprop224: Add test for parsing RENDEZVOUS1 cell.
George Kadianakis [Mon, 25 Sep 2017 12:00:06 +0000 (15:00 +0300)] 
prop224: Add test for parsing RENDEZVOUS1 cell.

Just making sure that added padding does not screw up cell parsing.

8 years agofix typo
Roger Dingledine [Fri, 22 Sep 2017 21:26:29 +0000 (17:26 -0400)] 
fix typo

8 years agoMention even more about changes files in doc/HACKING
Nick Mathewson [Fri, 22 Sep 2017 19:14:36 +0000 (15:14 -0400)] 
Mention even more about changes files in doc/HACKING

8 years agoFix check-spaces and a weird copyright.
Nick Mathewson [Fri, 22 Sep 2017 16:20:26 +0000 (12:20 -0400)] 
Fix check-spaces and a weird copyright.

8 years agoUse setup_full_capture_of_logs for 21800
Nick Mathewson [Fri, 22 Sep 2017 14:22:47 +0000 (10:22 -0400)] 
Use setup_full_capture_of_logs for 21800

setup_capture_of_logs() doesn't actually suppress warnings :p

8 years agoRemove unused chosen_scheduler_type variable.
Nick Mathewson [Fri, 22 Sep 2017 13:56:54 +0000 (09:56 -0400)] 
Remove unused chosen_scheduler_type variable.

8 years agoMerge remote-tracking branch 'pastly/bug23552_032_03'
Nick Mathewson [Fri, 22 Sep 2017 13:54:41 +0000 (09:54 -0400)] 
Merge remote-tracking branch 'pastly/bug23552_032_03'

8 years agoMerge branch 'maint-0.3.0' into maint-0.3.1
Nick Mathewson [Fri, 22 Sep 2017 13:09:17 +0000 (09:09 -0400)] 
Merge branch 'maint-0.3.0' into maint-0.3.1

8 years agoMerge branch 'maint-0.3.1'
Nick Mathewson [Fri, 22 Sep 2017 13:09:17 +0000 (09:09 -0400)] 
Merge branch 'maint-0.3.1'

8 years agohs: Avoid possible double circuit close on error
David Goulet [Thu, 21 Sep 2017 17:13:44 +0000 (13:13 -0400)] 
hs: Avoid possible double circuit close on error

Fixes #23610

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agoMerge branch 'bug22805_v2_squashed'
Nick Mathewson [Fri, 22 Sep 2017 12:58:17 +0000 (08:58 -0400)] 
Merge branch 'bug22805_v2_squashed'

8 years agoStop using CREATE_FAST as a signifier of clienthood.
Nick Mathewson [Fri, 15 Sep 2017 13:43:13 +0000 (09:43 -0400)] 
Stop using CREATE_FAST as a signifier of clienthood.

Relays send it, and we may as well let them.  Part of our fix for
22805.

8 years agochanges file for bug22805
Nick Mathewson [Wed, 6 Sep 2017 20:18:40 +0000 (16:18 -0400)] 
changes file for bug22805

8 years agoUse channel_is_client() accessor in channelpadding.c.
Nick Mathewson [Wed, 6 Sep 2017 19:37:09 +0000 (15:37 -0400)] 
Use channel_is_client() accessor in channelpadding.c.

Also, allow channel_is_client() to take a const channel.

8 years agoStop clearing the is_client flag on channel directly
Nick Mathewson [Wed, 6 Sep 2017 18:55:58 +0000 (14:55 -0400)] 
Stop clearing the is_client flag on channel directly

8 years agoRemove or_circuit_t.is_first_hop; use channel_is_client() instead
Nick Mathewson [Wed, 6 Sep 2017 18:39:40 +0000 (14:39 -0400)] 
Remove or_circuit_t.is_first_hop; use channel_is_client() instead

The is_first_hop field should have been called used_create_fast,
but everywhere that we wanted to check it, we should have been
checking channel_is_client() instead.

8 years agosched: move code to respect comments
Matt Traudt [Thu, 21 Sep 2017 17:24:07 +0000 (13:24 -0400)] 
sched: move code to respect comments

The diff is confusing, but were two static scheduler functions that
needed moving to static comment block.

No code change. Thanks dgoulet for original commit

8 years agosched: only log when scheduler type changes
Matt Traudt [Thu, 21 Sep 2017 17:15:51 +0000 (13:15 -0400)] 
sched: only log when scheduler type changes

Closes 23552. Thanks dgoulet for original impl

8 years agotest_util: capture the log in the correct place.
Nick Mathewson [Fri, 22 Sep 2017 12:51:03 +0000 (08:51 -0400)] 
test_util: capture the log in the correct place.

8 years agoMerge branch 'bug21800'
Nick Mathewson [Thu, 21 Sep 2017 21:39:54 +0000 (17:39 -0400)] 
Merge branch 'bug21800'

8 years agocheck-changes fix
Nick Mathewson [Thu, 21 Sep 2017 21:34:47 +0000 (17:34 -0400)] 
check-changes fix

8 years agoMerge remote-tracking branch 'pastly/bug23581_032_03'
Nick Mathewson [Thu, 21 Sep 2017 21:33:10 +0000 (17:33 -0400)] 
Merge remote-tracking branch 'pastly/bug23581_032_03'

8 years agoWhoops; only run lintchanges when usepython
Nick Mathewson [Thu, 21 Sep 2017 21:32:17 +0000 (17:32 -0400)] 
Whoops; only run lintchanges when usepython

8 years agoUpdate CodingStandards.md for changes-file stuff
Nick Mathewson [Thu, 21 Sep 2017 20:39:49 +0000 (16:39 -0400)] 
Update CodingStandards.md for changes-file stuff

'check-changes' is now part of "make check", and must pass.

8 years agoWhoops -- put a changes file in the wrong place :(
Nick Mathewson [Thu, 21 Sep 2017 20:29:15 +0000 (16:29 -0400)] 
Whoops -- put a changes file in the wrong place :(

8 years agoFix warning in bug23610
Nick Mathewson [Thu, 21 Sep 2017 20:27:09 +0000 (16:27 -0400)] 
Fix warning in bug23610

8 years agoMerge branch 'ticket23564_squashed'
Nick Mathewson [Thu, 21 Sep 2017 20:25:30 +0000 (16:25 -0400)] 
Merge branch 'ticket23564_squashed'

8 years agoMake check-spaces part of the standard "make check" process
Nick Mathewson [Tue, 19 Sep 2017 14:10:38 +0000 (10:10 -0400)] 
Make check-spaces part of the standard "make check" process

To do this, we had to make sure it passes when the changes directory
is empty.  I also tried to improve the quality of the output, and
fix some false-positive cases.  Let's see how this goes!

Closes ticket 23564.

8 years agoCapture "Result does not fit in tor_timegm()" warnings in util tests
Nick Mathewson [Thu, 21 Sep 2017 19:32:16 +0000 (15:32 -0400)] 
Capture "Result does not fit in tor_timegm()" warnings in util tests

Fixes bug 21800; bugfix on 0.2.9.x

8 years agoMerge remote-tracking branch 'dgoulet/ticket23303_032_01'
Nick Mathewson [Thu, 21 Sep 2017 18:43:05 +0000 (14:43 -0400)] 
Merge remote-tracking branch 'dgoulet/ticket23303_032_01'

8 years agoMerge remote-tracking branch 'catalyst-oniongit/bug23611'
Nick Mathewson [Thu, 21 Sep 2017 18:42:03 +0000 (14:42 -0400)] 
Merge remote-tracking branch 'catalyst-oniongit/bug23611'

8 years agoMerge branch 'maint-0.3.1'
Nick Mathewson [Thu, 21 Sep 2017 18:39:17 +0000 (14:39 -0400)] 
Merge branch 'maint-0.3.1'

8 years agohs: Avoid possible double circuit close on error
David Goulet [Thu, 21 Sep 2017 17:13:44 +0000 (13:13 -0400)] 
hs: Avoid possible double circuit close on error

Fixes #23610

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agoRemove wrong config parameter mention from manpage
Taylor Yu [Thu, 21 Sep 2017 18:31:54 +0000 (13:31 -0500)] 
Remove wrong config parameter mention from manpage

Don't refer to the obsolete (and misspelled)
UseEntryGuardsAsDirectoryGuards config parameter in the manpage.

8 years agohs: Improve comments for when we change timestamps
David Goulet [Thu, 21 Sep 2017 18:24:36 +0000 (14:24 -0400)] 
hs: Improve comments for when we change timestamps

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agosched: Hard exit if we can't select a scheduler
David Goulet [Tue, 19 Sep 2017 18:07:39 +0000 (14:07 -0400)] 
sched: Hard exit if we can't select a scheduler

Fixes #23581

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agohs: Avoid possible double circuit close on error
David Goulet [Thu, 21 Sep 2017 17:13:44 +0000 (13:13 -0400)] 
hs: Avoid possible double circuit close on error

Fixes #23610

Signed-off-by: David Goulet <dgoulet@torproject.org>
8 years agoMerge remote-tracking branch 'catalyst-oniongit/bug23607'
Nick Mathewson [Thu, 21 Sep 2017 14:58:06 +0000 (10:58 -0400)] 
Merge remote-tracking branch 'catalyst-oniongit/bug23607'

8 years agolintchanges fix
Nick Mathewson [Thu, 21 Sep 2017 14:42:40 +0000 (10:42 -0400)] 
lintchanges fix

8 years agoMerge branch 'maint-0.3.1'
Nick Mathewson [Thu, 21 Sep 2017 14:39:10 +0000 (10:39 -0400)] 
Merge branch 'maint-0.3.1'

8 years agoBug 23608: Mock time for all channelpadding tests.
Mike Perry [Wed, 20 Sep 2017 22:37:05 +0000 (22:37 +0000)] 
Bug 23608: Mock time for all channelpadding tests.

8 years agoAvoid assertion failure for state file clock skew
Taylor Yu [Wed, 20 Sep 2017 23:47:15 +0000 (18:47 -0500)] 
Avoid assertion failure for state file clock skew

The clock_skew_warning() refactoring allowed calls from
or_state_load() to control_event_bootstrap_problem() to occur prior
bootstrap phase 0, causing an assertion failure.  Initialize the
bootstrap status prior to calling clock_skew_warning() from
or_state_load().

8 years agoUse correct sign for state file clock skew
Taylor Yu [Wed, 20 Sep 2017 21:54:56 +0000 (16:54 -0500)] 
Use correct sign for state file clock skew

or_state_load() was using an incorrect sign convention when calling
clock_skew_warning() to warn about state file clock skew.  This caused
the wording of the warning to be incorrect about the direction of the
skew.

8 years agoWhen detecting OpenSSL, look for 1.0.1, and in homebrew location
Nick Mathewson [Wed, 20 Sep 2017 19:54:36 +0000 (15:54 -0400)] 
When detecting OpenSSL, look for 1.0.1, and in homebrew location

Previously we would detect the system openssl on OSX, and then fail
to use it, since we required Open 1.0.1 or later.  That's silly!

Instead of looking for RAND_add(), look for TLSv1_1_method(): it was
introduced in 1.0.1, and is also present in LibreSSL.

Also, add the hombebrew path to our search path here.

Fixes bug 23602; bugfix on 0.2.7.2-alpha.

8 years agoUse correct name for authtype 3 in a comment.
Nick Mathewson [Wed, 20 Sep 2017 19:46:21 +0000 (15:46 -0400)] 
Use correct name for authtype 3 in a comment.

8 years agohs: Log the intro point when we clean it up
David Goulet [Wed, 20 Sep 2017 16:47:09 +0000 (12:47 -0400)] 
hs: Log the intro point when we clean it up

When we remove an intro point from the service list, log info about it and
some useful data.

Signed-off-by: David Goulet <dgoulet@torproject.org>