Commit Graph

27092 Commits

Author SHA1 Message Date
ArunaMaurya221B
ad5cfa3039 Bug:24531 Function to change channel scheduler state for easy debugging added. 2017-12-11 08:46:43 -05:00
ArunaMaurya221B
d77cacb7ab Bug:24531 Add function to change scheduler state and always use it 2017-12-11 08:46:39 -05:00
Nick Mathewson
58e8094816 Fix compilation with --disable-memory-sentinels
We'd broken this with the recent _free() rewrite.
2017-12-11 08:01:54 -05:00
Nick Mathewson
241b676638 Fix up test_circuitstats to use the new circuit_free macro 2017-12-08 17:52:45 -05:00
Nick Mathewson
0340c02975 Merge remote-tracking branch 'mikeperry/bug23114_squashed2' 2017-12-08 17:50:34 -05:00
Nick Mathewson
b189c613bb move a macro; fix a build? 2017-12-08 15:11:18 -05:00
Nick Mathewson
5ee0cccd49 Merge branch 'macro_free_v2_squashed' 2017-12-08 14:58:43 -05:00
Nick Mathewson
7ca5f4bf03 document our allocator conventions 2017-12-08 14:47:19 -05:00
Nick Mathewson
fa0d24286b Convert remaining function (mostly static) to new free style 2017-12-08 14:47:19 -05:00
Nick Mathewson
17dcce3fe1 Fix wide lines introduced by previous patch. 2017-12-08 14:47:19 -05:00
Nick Mathewson
285632a61b Replace all FREE_AND_NULL* uses to take a type and a free function.
This commit was made mechanically by this perl script:

\#!/usr/bin/perl -w -i -p

next if /^#define FREE_AND_NULL/;
s/\bFREE_AND_NULL\((\w+),/FREE_AND_NULL\(${1}_t, ${1}_free_,/;
s/\bFREE_AND_NULL_UNMATCHED\(/FREE_AND_NULL\(/;
2017-12-08 14:47:19 -05:00
Nick Mathewson
95531ddfbf Let's have only one FREE_AND_NULL variant.
This commit removes the old FREE_AND_NULL, and renames the old
FREE_AND_NULL_UNMATCHED so that it is now called FREE_AND_NULL.

This will break all the FREE_AND_NULL_* users; the next commit will
fix them.
2017-12-08 14:47:19 -05:00
Nick Mathewson
5508ee29d5 changes file for big free macro-ization branch 2017-12-08 14:47:19 -05:00
Nick Mathewson
1d348989b0 Make tor_free only evaluate its input once (at least on gcc and clang) 2017-12-08 14:47:19 -05:00
Nick Mathewson
db024adc90 Switch to a safer FREE_AND_NULL implementation
This one only evaluates the input once, so it cannot mess up even if
there are side effects.
2017-12-08 14:47:19 -05:00
Nick Mathewson
176ad729d9 Change the free macro convention in the rest of src/or/*.h 2017-12-08 14:47:19 -05:00
Nick Mathewson
0792cc107e Convert connection_free to a nulling macro. 2017-12-08 14:47:19 -05:00
Nick Mathewson
a48ba072a9 Rename connection_free_ to connection_free_minimal. 2017-12-08 14:47:19 -05:00
Nick Mathewson
b0cc9856ee Update free functions into macros: src/or/ part 1
This covers addressmap.h (no change needed) through confparse.h
2017-12-08 14:47:19 -05:00
Nick Mathewson
c92ac9f5cb Convert the rest of src/common's headers to use FREE_AND_NULL 2017-12-08 14:47:19 -05:00
Nick Mathewson
44010c6fc1 Merge branch 'dgoulet_ticket23709_033_01_squashed' 2017-12-08 14:44:09 -05:00
David Goulet
50124fd7a6 chan: Add changes file for ticket 23709
Closes #23709

Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-12-08 14:43:27 -05:00
David Goulet
e96c577ed2 test: Make older GCC happy and thus our oniongit pipeline
Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-12-08 14:43:27 -05:00
David Goulet
6120efd771 chan: Do not re-queue after a fail cell write
Couple things happen in this commit. First, we do not re-queue a cell back in
the circuit queue if the write packed cell failed. Currently, it is close to
impossible to have it failed but just in case, the channel is mark as closed
and we move on.

The second thing is that the channel_write_packed_cell() always took ownership
of the cell whatever the outcome. This means, on success or failure, it needs
to free it.

It turns out that that we were using the wrong free function in one case and
not freeing it in an other possible code path. So, this commit makes sure we
only free it in one place that is at the very end of
channel_write_packed_cell() which is the top layer of the channel abstraction.
This makes also channel_tls_write_packed_cell_method() return a negative value
on error.

Two unit tests had to be fixed (quite trivial) due to a double free of the
packed cell in the test since now we do free it in all cases correctly.

Part of #23709

Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-12-08 14:43:27 -05:00
teor
bb7c5b431d
Move a comment to relay_send_end_cell_from_edge()
It looks like it was left behind in a refactor.

Fixes 24559.
2017-12-08 12:49:33 +11:00
Nick Mathewson
afceb431ed add a missing windows underscore 2017-12-07 15:14:49 -05:00
Nick Mathewson
45b9b2245f Merge branch 'maint-0.3.2' 2017-12-07 08:42:56 -05:00
Nick Mathewson
4b41ffa64d Merge branch 'arthuredelstein_18859+1_031' into maint-0.3.2 2017-12-07 08:42:49 -05:00
Nick Mathewson
13049a9866 Rewrite 18859 changes file from user POV. 2017-12-07 08:41:00 -05:00
Nick Mathewson
c2c0f83c23 Revert accidentally-committed code from 046acf208b 2017-12-07 08:28:57 -05:00
Mike Perry
050bb67974 Add tests for circuitstats.c
These tests primarily test the relaxed and measured behavior of
circuitstats.c, to make sure we did not break it with #23100 or #23114.
2017-12-07 00:04:39 +00:00
Mike Perry
125df07d60 Report close and timeout rates since uptime, not based on data.
Bug #23114 was harder to see because we were just reporting our math,
rather than reporting behavior.
2017-12-07 00:04:39 +00:00
Mike Perry
ed89588c4f Bug #23114: Time out circuits immediately.
This changes the purpose of circuits that are past the timeout to measurement
*as they are built*, ensuring accurate application of the timeout logic.
2017-12-07 00:04:38 +00:00
Mike Perry
b5d4cd1b41 Bug #23100: Count all 3 hop circuits for CBT.
This change causes us to count anything once it reaches 3 hops (but not
after).
2017-12-07 00:04:33 +00:00
Arthur Edelstein
2bd91dbd34 Don't consider a port "handled" by an isolated circuit.
Previously, circuit_stream_is_being_handled incorrectly reported
that (1) an exit port was "handled" by a circuit regardless of
whether the circuit was already isolated in some way, and
(2) that a stream could be "handled" by a circuit even if their
isolation settings were incompatible.

As a result of (1), in Tor Browser, circuit_get_unhandled_ports was
reporting that all ports were handled even though all non-internal
circuits had already been isolated by a SOCKS username+password.

Therefore, circuit_predict_and_launch_new was declining to launch
new exit circuits. Then, when the user visited a new site in Tor
Browser, a stream with new SOCKS credentials would be initiated,
and the stream would have to wait while a new circuit with those
credentials could be built. That wait was making the
time-to-first-byte longer than it needed to be.

Now, clean, not-yet-isolated circuit(s) will be automatically
launched ahead of time and be ready for use whenever a new stream
with new SOCKS credentials (or other isolation criteria) is
initiated.

Fixes bug 18859. Thanks to Nick Mathewson for improvements.
2017-12-06 14:18:41 -08:00
Nick Mathewson
046acf208b Fix a compiler warning 2017-12-06 15:46:54 -05:00
Nick Mathewson
5f518c69aa Merge remote-tracking branch 'public/monotime_coarse_stamps' 2017-12-06 15:43:50 -05:00
Nick Mathewson
c470a6d278 Merge remote-tracking branch 'teor/bug24488' 2017-12-06 14:44:03 -05:00
Nick Mathewson
1432cc8429 Merge branch 'maint-0.3.2' 2017-12-06 14:38:06 -05:00
Nick Mathewson
d4ca18573c Merge remote-tracking branch 'dgoulet/bug24502_032_01' into maint-0.3.2 2017-12-06 14:37:53 -05:00
Nick Mathewson
91f72bbb6c Merge remote-tracking branch 'teor/bug24489' 2017-12-06 14:37:15 -05:00
Nick Mathewson
aa6212a858 Merge remote-tracking branch 'dgoulet/bug24502_032_01' 2017-12-06 14:31:33 -05:00
David Goulet
1a55a5ff06 test: Add a KIST test for a non opened channel
This makes sure that a non opened channel is never put back in the channel
pending list and that its state is consistent with what we expect that is
IDLE.

Test the fixes in #24502.

Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-12-06 11:33:01 -05:00
Alexander Færøy
9fa38bd996
Check the return value of hs_parse_address().
This patch adds a check for the return value of `hs_parse_address()` in
`hs_control_hspost_command()`. Since it should not be possible for
`hs_parse_address()` to fail in this context we wrap the error check
with the `BUG()` macro.

See: https://bugs.torproject.org/24543
2017-12-06 14:56:52 +01:00
Nick Mathewson
01700d779c Initialize pk so that older gcc versions don't freak out. 2017-12-05 20:00:49 -05:00
Nick Mathewson
a7a0cebb59 Merge branch 'more_directories_squashed' 2017-12-05 19:49:45 -05:00
Nick Mathewson
46f167886a add a changes file 2017-12-05 19:49:29 -05:00
Nick Mathewson
02cbf2ffc1 Update the manpage to describe {Cache,Key}Directory
Also, explain which files should be put in which.
2017-12-05 19:49:29 -05:00
Nick Mathewson
a57bcffcc7 Implement the various get_foodir_*() functions. 2017-12-05 19:49:29 -05:00
Nick Mathewson
a9806af261 Create a CacheDirectory and KeyDirectory options.
They work the same as DataDirectory, but default slightly different.

Tor is not actually updated to use them yet.
2017-12-05 19:49:28 -05:00