When ProcessMessageBatchedSigShares returns false, it's interpreted as
if an invalid/malicious message was received, causing a ban. So, we should
return "!ban" instead of just "ban".
It was reported on iOS that CMerkleBlock sometimes included the dummy
quorum commitments introduced with v13, which led to banning of nodes as
these were not supported/expected there.
We should in general only include TXs here that are of interest for SPV
nodes, so we should maintain the list of allowed TX types.
* Ignore sig share inv messages when we don't have the quorum vvec
* Update src/llmq/quorums_signing_shares.cpp
Co-Authored-By: codablock <ablock84@gmail.com>
We may eventually want to break out harfbuzz and build it in depends, but
for now just ensure that runtime dependencies don't depend on whether or not
harfbuzz was present on the builder.
* On timeout, print members proTxHashes from members which did not send a share
* Move inactive quorums check above timeout checks
This allows to reuse things in the next commit
* Avoid locking cs_main through GetQuorum by using a pre-filled map
* Use find() instead of [] to access quorums map
* Return bool in ProcessMessageXXX methods to indicate misbehaviour
* Send/Receive multiple messages as part of one P2P message in CSigSharesManager
Many messages, especially QSIGSHARESINV and QGETSIGSHARES, are very small
by nature (5-14 bytes for a 50 members LLMQ). The message headers are
24 bytes, meaning that we produce a lot of overhead for these small messages.
This sums up quite a bit when thousands of signing sessions are happening
in parallel.
This commit changes all related P2P messages to send a vector of messages
instead of a single message.
* Remove bogus lines
Included these by accident
* Unify handling of BanNode in ProcessMessageXXX methods
* Remove bogus check for fMasternodeMode
* Properly use == instead of misleading >= in SendMessages
* Put "didSend = true" near PushMessage
Stop relying on the information previously found in the CSigSharesInv
and CBatchedSigShares messages and instead use the information found in
the session refereced by the session id.
This also updates a few LogPrintf calls. Previously, CSigSharesInv::ToString
also included the signHash in the returned string, which is not the case
anymore, so we have to add it manually.
We must watch out to not blindly use externally provided keys in unordered
sets/maps, as attackers might find ways to cause unbalanced hash buckets
causing performance degradation.
Allows convenient salted hashing with unordered maps and sets. Useful when
there is a risk of unbalanced hash buckets slowing things down, e.g. when
externally supplied hashes are used as keys into a map.
* Indicate success when signing was unnecessary
* Fix typo in name of LLMQ_400_60
* Move RemoveAskFor call for CLSIGs into ProcessNewChainLock
In case we got INV items for the same CLSIG that we recreated through
HandleNewRecoveredSig, (re-)requesting of the CLSIG from other peers
becomes unnecessary.
* Move Cleanup() call in CChainLocksHandler::UpdatedBlockTip up
We bail out early in a few situations from this method, so that Cleanup()
might not be called while its at the bottom.
* Bail out from CChainLocksHandler::UpdatedBlockTip if we already got the CLSIG
* Call RemoveAskFor when QFCOMMITMENT was received
Otherwise we might end up re-requesting it for a very long time when the
commitment INV was received shortly before it got mined.
* Call RemoveSigSharesForSession when a recovered sig is received
Otherwise we end up with session data in node states lingering around until
a fake "timeout" occurs (can be seen in the logs).
* Better handling of false-positive conflicts in CSigningManager
The old code was emitting a lot of messages in logs as it treated sigs
for exactly the same session as a conflict. This commit fixes this by
looking at the signHash before logging.
Also handle a corner-case where a recovered sig might be deleted between
the HasRecoveredSigForId and GetRecoveredSigById call.
* Don't run into session timeout when sig shares come in slow
Instead of just tracking when the first share was received, we now also
track when the last (non-duplicate) share was received. Sessios will now
timeout 5 minutes after the first share arrives, or 1 minute after the last
one arrived.
* Add libbacktrace to depends
This is currently only useful to extract symbols. It fails to gather
stacktraces when compiled with MinGW, so we can only use it to get symbol
information from a stack trace which we gathered outside of libbacktrace.
* Add -mbig-obj to CXXFLAGS for MinGW builds
* Implement stacktraces for C++ exceptions
This is a hack and should only be used for debugging. It works by wrapping
the C++ ABI __wrap___cxa_allocate_exception. The wrapper records a backtrace
and stores it in a global map. Later the stacktrace can be retrieved with
GetExceptionStacktraceStr.
This commit also adds handlers to pretty print uncaught exceptions and
signals.
* Use GetPrettyExceptionStr for all unhandled exceptions
* Use --enable-stacktraces in CI for linux32/linux64
* Register exception translators to pretty print exceptions in unit tests
* Catch and print python exceptions when stopping nodes
Otherwise the code at the bottom is never executed when nodes crash,
leading to no output of debug.log files on Travis.
* Remove now unneeded/unused TestCrash methods
Instead of trying to manually figure out params for different quorum/ring sizes, connect to nodes at indexes (i+2^k)%n where k: 0..floor(log2(n-1))-1, n: size of the quorum/ring
* Implement and use SigShareMap instead of ordered map with helper methods
The old implementation was relying on the maps being ordered, which allowed
us to grab all sig shares for the same signHash by doing range queries on
the map. This has the disadvantage of being unnecessarily slow when the
maps get larger. Using an unordered map would be the naive solution, but
then it's not possible to query by range anymore.
The solution now is to have a specialized map "SigShareMap" which is
indexed by "SigShareKey". It's internally just an unordered map, indexed by
the sign hash and another unordered map for the value, indexed by the
quorum member index.
* Only use unordered maps/sets in CSigSharesManager
These are faster when maps/sets get larger.
* Use unorderes sets/maps in CSigningManager
* Don't sleep in WorkThreadMain when CPU intensive work was done
When the current iteration resulted in CPU intensive work, it's likely that
the next iteration will result in work as well. Do not sleep in that case,
as we're otherwise wasting (unused) CPU resources.
* No matter how fast we process sig shares, always force 100ms between sending
* Apply review suggestions
This removes the burden on the message handler thread when many sig batches
arrive. The expensive part of deserialization is now performed in the sig
shares worker thread.
This also removes the need for the specialized deserialization of the sig
shares which tried to avoid the malleability check, as CBLSLazySignature does
not perform malleability checks at all.
In some cases it takes too much time to perform full deserialization of
BLS signatures in the message handler thread. Better to just read the
buffer and do the actual deserialization when the signature is needed for
the first time (which is can be in another thread).
This adds the reading side of a pipe to the read-set when calling select().
Writing to the writing side of the pipe then causes select() to wake up
immediately. Otherwise it would wait for the timeout of 50ms, even if there
is data that could possibly be sent.
This is useful when many messages need are pushed with optimistic send being
disabled. After all messages have been pushed, WakeSelect() can then wakeup
the select() thread and force a re-check for pending data to send.
This is currently only implemented for POSIX compliant systems as we assume
that heavy-load daemons (like masternodes) are usually run on Linux.
Profiling has shown that optimistic send causes measurable slowdowns when
many messages are pushed, even if the sockets are non-blocking. Better to
allow disabling of optimistic sending in such cases and let the network
thread do the actual socket calls.
