This is a _subtle_ bug introduced by d1494d14, which resolved
connections that was allocated in the extorport/handshake test. So
how did the connection get freed? Our test was set up so that every
extorport connection would get the same ext_or_id. Two connections
couldn't have the same ext_or_id, and if they did, one would get
freed. This meant that the _next_ connection to be constructed in
the test would cause the previous connection to become closeable,
even if it hadn't been closeable before.
But when we applied d149d14, we stopped making it so our code
enforced this uniqueness, and thereby make it so we _weren't_
freeing this connection in the tests.
Closes#40260; bug not in any released version of Tor.
This step happens after we make each consensus flavor, and before we
worry about sigs or anything. That way if Tor crashes, or if we fail to
get enough sigs, we still have a chance to know what consensus we wanted
to make.
This validation was only done if DisableNetwork was off because we would use
the global list of transports/bridges and DisableNetwork would not populate
it.
This was a problem for any user using DisableNetwork which includes Tor
Browser and thus leading to the Bug() warning.
Without a more in depth refactoring, we can't do this validation without the
global list.
The previous commit makes it that any connection to a bridge without a
transport won't happen thus we keep the security feature of not connecting to
a bridge without its corresponding transport.
Related to #40106
Signed-off-by: David Goulet <dgoulet@torproject.org>
Don't pick the bridge as the guard or launch descriptor fetch if no transport
is found.
Fixes#40106
Signed-off-by: David Goulet <dgoulet@torproject.org>
This patch limits the number of items in the consensus diff cache to 64
on the Windows platform. Hopefully, this will allow us to investigate a
smarter fix while avoiding the situation reported in tor#24857 where
Windows relay operators report Tor using 100% CPU.
See: tor#24857
When selecting the first advertised port, we always prefer the one with an
explicit address.
Closes#40246
Signed-off-by: David Goulet <dgoulet@torproject.org>
We used to actually discard ORPorts that were the same port and same family
but they could have different address.
Instead, we need to keep all different ORPorts so we can bind a listener on
each of them. We will publish only one of these in our descriptor though.
Related to #40246
Signed-off-by: David Goulet <dgoulet@torproject.org>
This patch removes a call to `tor_assert_nonfatal()` if
`extend_info_from_node()` returns NULL. This is unnecessary as we
already handle the case where `info` is NULL in the next `if (!info) {
... }` block in the code.
See: tor#32666.
This reverts commit d07f17f676.
We don't want to consider an entire routable IPv6 network as sybil if more
than 2 relays happen to be on it. For path selection it is very important but
not for selecting relays in the consensus.
Fixes#40243
We can end up trying to find our address from an authority while we don't have
yet its descriptor.
In this case, don't BUG() and just come back later.
Closes#40231
Signed-off-by: David Goulet <dgoulet@torproject.org>
In case building the descriptor would fail, we could still flag that we did in
fact publish the descriptors leading to no more attempt at publishing it which
in turn makes the relay silent for some hours and not try to rebuild the
descriptor later.
This has been spotted with #40231 because the operator used a localhost
address for the ORPort and "AssumeReachable 1" leading to this code path where
the descriptor failed to build but all conditions to "can I publish" were met.
Related to #40231
Signed-off-by: David Goulet <dgoulet@torproject.org>
This one should work on GCC _and_ on Clang. The previous version
made Clang happier by not having unreachable "fallthrough"
statements, but made GCC sad because GCC didn't think that the
unconditional failures were really unconditional, and therefore
_wanted_ a FALLTHROUGH.
This patch adds a FALLTHROUGH_UNLESS_ALL_BUGS_ARE_FATAL macro that
seems to please both GCC and Clang in this case: ordinarily it is a
FALLTHROUGH, but when ALL_BUGS_ARE_FATAL is defined, it's an
abort().
Fixes bug 40241 again. Bugfix on earlier fix for 40241, which was
merged into maint-0.3.5 and forward, and released in 0.4.5.3-rc.
Our original code for parsing these parameters out of our list of
parameters pre-dated us having the
dirvote_get_intermediate_param_value() function... and it was buggy.
Specifically, it would reject any " ... K=V ..." value
where there were additional unconverted characters after the V, and
use the default value instead,
We haven't run into this yet because we've never voted for
bwweightscale to be anything besides the default 10000, or
maxunmeasuredbw to be anything besides the default 20.
This requires a new consensus method because it is a change in how
consensuses are computed.
Fixes bug 19011; bugfix on 0.2.2.10-alpha.
Some days before this commit, the network experienced a DDoS on the directory
authorities that prevented them to generate a consensus for more than 5 hours
straight.
That in turn entirely disabled onion service v3, client and service side, due
to the subsystem requiring a live consensus to function properly.
We know require a reasonably live consensus which means that the HSv3
subsystem will to its job for using the best consensus tor can find. If the
entire network is using an old consensus, than this should be alright.
If the service happens to use a live consensus while a client is not, it
should still work because the client will use the current SRV it sees which
might be the previous SRV for the service for which it still publish
descriptors for.
If the service is using an old one and somehow can't get a new one while
clients are on a new one, then reachability issues might arise. However, this
is a situation we already have at the moment since the service will simply not
work if it doesn't have a live consensus while a client has one.
Fixes#40237
Signed-off-by: David Goulet <dgoulet@torproject.org>
Thanks to proposal 315 / ticket #30132, more fields are now
required in these documents. But ancient Tors that try to upload
obsolete documents were causing the authorities to log warnings
about missing fields, and to do so very spammily.
We now detect the missing fields before tokenizing, and log at
debug. This is a bit of ugliness, but it's probably a safer choice
than making _all_ unparseable-desc warnings into debug-level logs.
I'm looking at identity-ed25519 in extrainfos and proto in
routerdescs because they were (I believe) the latest-added fields in
Tor's history: any Tor that lacks them will also lack the other
newly required fields.
Fixes bug #40238; bugfix on 0.4.5.1-alpha.
