Commit Graph

26502 Commits

Author SHA1 Message Date
Nick Mathewson
ca2e7eabc2 Merge branches 'bug8185_031' and 'bug8185_diagnostic_032' 2017-09-29 17:26:40 -04:00
David Goulet
070064de89 sched: Always initialize scheduler_last_run to now
Because our monotonic time interface doesn't play well with value set to 0,
always initialize to now() the scheduler_last_run at init() of the KIST
scheduler.

Fixes #23696

Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-09-29 14:07:55 -04:00
Matt Traudt
3ef7e6f187 sched: Don't get KIST stuck in an infinite loop
When a channel is scheduled and flush cells returns 0 that is no cells to
flush, we flag it back in waiting for cells so it doesn't get stuck in a
possible infinite loop.

It has been observed on moria1 where a closed channel end up in the scheduler
where the flush process returned 0 cells but it was ultimately kept in the
scheduling loop forever. We suspect that this is due to a more deeper problem
in tor where the channel_more_to_flush() is actually looking at the wrong
queue and was returning 1 for an empty channel thus putting the channel in the
"Case 4" of the scheduler which is to go back in pending state thus
re-considered at the next iteration.

This is a fix that allows the KIST scheduler to recover properly from a not
entirelly diagnosed problem in tor.

Fixes #23676

Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-09-29 11:06:31 -04:00
Nick Mathewson
fc6c0b46fb Merge branch 'bug23690_additional_032' 2017-09-29 10:29:40 -04:00
Nick Mathewson
b0ddaac074 Make some assertions nonfatal to help prevent bug23690 recurrence. 2017-09-29 10:29:33 -04:00
Nick Mathewson
8ccefa7b7a Merge branch 'bug23690_028' 2017-09-29 10:28:33 -04:00
Nick Mathewson
f7222e6d8c Clear outbuf_flushlen when we clear a connection's outbuf
When we added single_conn_free_bytes(), we cleared the outbuf on a
connection without setting outbuf_flushlen() to 0.  This could cause
an assertion failure later on in flush_buf().

Fixes bug 23690; bugfix on 0.2.6.1-alpha.
2017-09-29 10:00:14 -04:00
Nick Mathewson
1a5746f8e1 Merge branch 'bug8185_025' into bug8185_031 2017-09-29 09:37:01 -04:00
Nick Mathewson
d256d4c0a6 Don't package cells onto marked circuits.
This caused a BUG log when we noticed that the circuit had no
channel.  The likeliest culprit for exposing that behavior is
d769cab3e5, where we made circuit_mark_for_close() NULL out
the n_chan and p_chan fields of the circuit.

Fixes bug 8185; bugfix on 0.2.5.4-alpha, I think.
2017-09-29 09:33:47 -04:00
Nick Mathewson
1b6ccb52ea Log more information when bug8185 is about to trigger.
My current theory is that this is just a marked circuit that hasn't
closed yet, but let's gather more information in case that theory is
wrong.

Diagnostic for 8185.
2017-09-29 09:24:01 -04:00
Isis Lovecruft
40f1287978
doc: Add a good resource for learning to write Rust FFI. 2017-09-29 01:04:18 +00:00
Nick Mathewson
4e6374c26b Merge remote-tracking branch 'catalyst-oniongit/bug23691' 2017-09-28 19:07:23 -04:00
Nick Mathewson
fba30b8b26 Merge remote-tracking branch 'ahf-gitlab/bugs/23551' 2017-09-28 18:54:40 -04:00
Nick Mathewson
4110745017 Merge branch 'maint-0.3.1' 2017-09-28 18:54:18 -04:00
Taylor Yu
2ae1b7e2db Fix leak in test_util.c
AddressSanitizer found a leak in test_util_decompress_dos_impl().
Fixes bug #23691.
2017-09-28 17:53:38 -05:00
Nick Mathewson
82f109c2b3 Merge remote-tracking branch 'ahf-gitlab/bugs/23551-maint-0.3.1' into maint-0.3.1 2017-09-28 18:53:37 -04:00
Alexander Færøy
8d6940814a
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
2017-09-28 20:17:41 +02:00
Alexander Færøy
c2fac2c6b0
Enable disabled test that was disabled due to bug #23551
See: https://bugs.torproject.org/23551
2017-09-28 18:58:15 +02:00
Alexander Færøy
44dc4b73ec
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
2017-09-28 18:58:15 +02:00
Alexander Færøy
c3b7f9d762
Fix whitespace issue in compress.c 2017-09-28 18:58:15 +02:00
Alexander Færøy
a196fdb622
Fix typo in buffers.c. 2017-09-28 18:58:09 +02:00
Nick Mathewson
3a073c463d 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.
2017-09-28 12:20:02 -04:00
Nick Mathewson
14614a592e Fix spelling: compressing, not compresing 2017-09-28 12:17:34 -04:00
Nick Mathewson
f71ff0cabc Unit test for case where %included dir exists but is unreadable 2017-09-28 09:49:30 -04:00
Nick Mathewson
5b6f2713cd unit test for config_lines_dup_and_filter 2017-09-28 09:40:17 -04:00
Nick Mathewson
55873107d0 Correct docs for config_lines_dup_and_filter 2017-09-28 09:40:04 -04:00
Nick Mathewson
a64d79ca4c 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.
2017-09-28 09:25:17 -04:00
Nick Mathewson
9727a9248a Improve cov-diff script to account for new files. 2017-09-28 08:53:16 -04:00
Nick Mathewson
99e52711bb Add unit test for buf_get_line(). 2017-09-28 08:35:30 -04:00
Nick Mathewson
01d67a9071 Note an unreachable (?) section in buffers.c 2017-09-28 08:35:24 -04:00
Nick Mathewson
f21102c069 Let's get 100% coverage for proto_http while we're at it.
There was some coverage before, but it was mostly accidental.
2017-09-27 21:45:55 -04:00
Nick Mathewson
57f04a482e Test more error cases of our socks code.
Coverage is now respectable. :)
2017-09-27 19:37:36 -04:00
Nick Mathewson
6882e711d0 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 :/
2017-09-27 18:55:14 -04:00
Nick Mathewson
5418aa84ee socks5 unit tests: add check for ipv6 address type. 2017-09-27 11:23:09 -04:00
Nick Mathewson
04f7873bab Socks tests for bad socks5 username/passwd auth. 2017-09-27 11:16:29 -04:00
Nick Mathewson
14124f82df 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.
2017-09-27 11:16:20 -04:00
Nick Mathewson
5427365907 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.
2017-09-27 10:54:56 -04:00
David Goulet
cf75c1af66 hs-v3: Don't fetch descriptor if we have a pending request
If 6 SOCKS requests are opened at once, it would have triggered 6 fetches
which ultimately poke all 6 HSDir. We don't want that, if we have multiple
SOCKS requests for the same service, do one fetch only.

Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-09-27 09:24:35 -04:00
George Kadianakis
178a2bbcff prop224: Fix one last instance of get_time_period_num() with now.
When purging last HSDir requests, we used time(NULL) for computing the
service blinded key but in all other places in our codebase we actually
use the consensus times. That can cause wrong behavior if the consensus
is in a different time period than time(NULL).

This commit is required for proper purging of HSDir requests.
2017-09-27 09:24:35 -04:00
George Kadianakis
6051375cca prop224: Update some function docs with the right retval. 2017-09-27 09:24:35 -04:00
Nick Mathewson
a57f495c1a Move ext_or_command tests to test_proto_misc.c
No other code changes.
2017-09-27 09:11:15 -04:00
Nick Mathewson
b8b016940d Add unit tests for var_cell and control0 proto functions. 2017-09-27 09:09:00 -04:00
Taylor Yu
8f8a4c0ced Eat redundant semicolons 2017-09-26 15:12:34 -04:00
Nick Mathewson
b8b4143517 fix the changes file 2017-09-26 14:51:35 -04:00
David Goulet
40ab25db62 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>
2017-09-26 13:55:51 -04:00
Nick Mathewson
2de588ffdb Merge branch 'ticket23602_029' 2017-09-26 13:02:26 -04:00
Nick Mathewson
6beeb10070 Merge branch 'typecheck4' 2017-09-26 12:26:01 -04:00
Nick Mathewson
8f0dffe329 changes file for my confparse typechecking fun 2017-09-26 12:25:34 -04:00
Nick Mathewson
49d2346b80 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.
2017-09-26 12:25:34 -04:00
Nick Mathewson
a5b18dfba9 Make the TransProxyType field non-const
The correct type for a STRING confparse value is char *, not const
char *.
2017-09-26 12:25:07 -04:00