Client & HS make sure every hop in every non-HS path supports ntor
When a client connects to an intro point not in the client's consensus,
or a hidden service connects to a rend point not in the hidden service's
consensus, we are stuck with using TAP, because there is no ntor link
specifier.
If we know a node's version, and it can't do ntor, consider it not running.
If we have a node's descriptor, and it doesn't have a valid ntor key,
consider it not running.
Refactor these checks so they're consistent between authorities and clients.
Before, they checked for version 0.2.4.18-rc or later, but this
would not catch relays without version lines, or buggy or malicious
relays missing an ntor key.
cypherpunks [Thu, 16 Jun 2016 12:16:57 +0000 (12:16 +0000)]
Fix test-network-all target in out-of-tree builds
The test-network-all target assumes the test-driver script lives in the
current working directory. This assumption breaks out-of-tree builds
because it actually lives in the source directory.
Automake 1.12 introduces `LOG_DRIVER` which defines the location of the
test driver script. Because Tor still supports Automake 1.11 we use the
default value of this variable directly. The default value uses the
configured shell for calling the test driver script and explicitly
prefixes the source directory.
When checking if a routerstatus is reachable, don't check the node
This fixes #19608, allowing IPv6-only clients to use
microdescriptors, while preserving the ability of bridge clients
to have some IPv4 bridges and some IPv6 bridges.
David Goulet [Tue, 5 Jul 2016 17:54:37 +0000 (13:54 -0400)]
Test: fix shared random test checking bad errno
The test was checking for EISDIR which is a Linux-ism making other OSes
unhappy. Instead of checking for a negative specific errno value, just make
sure it's negative indicating an error. We don't need more for this test.
Signed-off-by: David Goulet <dgoulet@torproject.org>
Run asciidoc in UTC timezone for build reproducibility.
asciidoc adds a timestamp at the end of a generated HTML file.
This timestamp is based on the date of the file but it can change
depending on the TZ environment variable.
Nick Mathewson [Tue, 5 Jul 2016 16:10:12 +0000 (12:10 -0400)]
Grammar.
I grepped and hand-inspected the "it's" instances, to see if any
were supposed to be possessive. While doing that, I found a
"the the", so I grepped to see if there were any more.
David Goulet [Mon, 4 Jul 2016 16:05:48 +0000 (12:05 -0400)]
sr: add the base16 RSA identity digest to commit
Keep the base16 representation of the RSA identity digest in the commit object
so we can use it without using hex_str() or dynamically encoding it everytime
we need it. It's used extensively in the logs for instance.
Fixes #19561
Signed-off-by: David Goulet <dgoulet@torproject.org>
Sebastian Hahn [Sun, 3 Jul 2016 17:42:36 +0000 (19:42 +0200)]
Raise libevent dependency to 2.0.10-stable or newer
Only some very ancient distributions don't ship with Libevent 2 anymore,
even the oldest supported Ubuntu LTS version has it. This allows us to
get rid of a lot of compat code.
Peter Palfrader [Sun, 3 Jul 2016 16:03:26 +0000 (18:03 +0200)]
sandboxing: allow writing to stats/hidserv-stats
Our sandboxing code would not allow us to write to stats/hidserv-stats,
causing tor to abort while trying to write stats. This was previously
masked by bug#19556.
Peter Palfrader [Sun, 3 Jul 2016 15:47:45 +0000 (17:47 +0200)]
sandboxing: allow open() of stats dir
When sandboxing is enabled, we could not write any stats to disk.
check_or_create_data_subdir("stats"), which prepares the private stats
directory, calls check_private_dir(), which also opens and not just stats() the
directory. Therefore, we need to also allow open() for the stats dir in our
sandboxing setup.
The test_state_update() test would fail if you run it between 23:30 and
00:00UTC in the following line because n_protocol_runs was 2:
tt_u64_op(state->n_protocol_runs, ==, 1);
The problem is that when you launch the test at 23:30UTC (reveal phase),
sr_state_update() gets called from sr_state_init() and it will prepare
the state for the voting round at 00:00UTC (commit phase). Since we
transition from reveal to commit phase, this would trigger a phase
transition and increment the n_protocol_runs counter.
The solution is to initialize the n_protocol_runs to 0 explicitly in the
beginning of the test, as we do for n_reveal_rounds, n_commit_rounds etc.
Nick Mathewson [Tue, 10 May 2016 00:10:43 +0000 (20:10 -0400)]
Refactor parameter computation and add a helper function
This patch makes us retain the intermediate list of K=V entries for
the duration of computing our vote, and lets us use that list with
a new function in order to look up parameters before the consensus
is published.
We can't actually use this function yet because of #19011: our
existing code to do this doesn't actually work, and we'll need a new
consensus method to start using it.
prop250: Don't use {0} to init static struct -- causes warning on clang.
See ticket #19132 for the clang/llvm warning.
Since voting_schedule is a global static struct, it will be initialized
to zero even without explicitly initializing it with {0}.
This is what the C spec says:
If an object that has automatic storage duration is not initialized
explicitly, its value is indeterminate. If an object that has static
storage duration is not initialized explicitly, then:
— if it has pointer type, it is initialized to a null pointer;
— if it has arithmetic type, it is initialized to (positive or unsigned) zero;
— if it is an aggregate, every member is initialized (recursively) according to these rules;
— if it is a union, the first named member is initialized (recursively) according to these rules.
David Goulet [Thu, 12 May 2016 17:30:07 +0000 (13:30 -0400)]
prop250: Fix unit tests about the RSA fingerprint check
Code has been changed so every RSA fingerprint for a commit in our state is
validated before being used. This fixes the unit tests by mocking one of the
key function and updating the hardcoded state string.
Also, fix a time parsing overflow on platforms with 32bit time_t
Signed-off-by: David Goulet <dgoulet@torproject.org> Signed-off-by: George Kadianakis <desnacked@riseup.net>
David Goulet [Tue, 3 May 2016 15:36:09 +0000 (11:36 -0400)]
prop250: Parse votes and consensus
One of the last piece that parses the votes and consensus in order to update
our state and make decision for the SR values.
We need to inform the SR subsystem when we set the current consensus because
this can be called when loaded from file or downloaded from other authorities
or computed.
The voting schedule is used for the SR timings since we are bound to the
voting system.
Signed-off-by: David Goulet <dgoulet@torproject.org> Signed-off-by: George Kadianakis <desnacked@riseup.net>
David Goulet [Tue, 3 May 2016 14:57:49 +0000 (10:57 -0400)]
prop250: Add commit and SR values generation code
This adds the logic of commit and SR values generation. Furthermore, the
concept of a protocol run is added that is commit is generated at the right
time as well as SR values which are also rotated before a new protocol run.
Signed-off-by: George Kadianakis <desnacked@riseup.net> Signed-off-by: David Goulet <dgoulet@torproject.org>
projects / thirdparty / tor.git / log
