Rather than having a separate query channel for the block manager, use the
same channel so the block handler acts as a pure FIFO queue. This
prevents possible starvation of query related messages.
ok @owainga
This commit makes a slight variant of the existing checkConnectBlock
function available to external callers. The new exported version checks
if the passed block will connect to the end of the current main chain.
In order to support this, a few other small modifications have been made
to the initial index generation and the existing checkConnectBlock since
it previously made some assumptions about the state of genesis block which
can no longer be assumed due accepting blocks from callers directly.
Also, add a quick test to ensure the new function fails when checking if
the genesis block will connect again when it's already inserted.
This change modifies the RPC server's notifiation manager from a
struct with requests, protected by a mutux, to two goroutines. The
first maintains a queue of all notifications and control requests
(registering/unregistering notifications), while the second reads from
the queue and processes notifications and requests one at a time.
Previously, to prevent slowing down block and mempool processing, each
notification would be handled by spawning a new goroutine. This lead
to cases where notifications would end up being sent to clients in a
different order than they were created. Adding a queue keeps the
order of notifications originating from the same goroutine, while also
not slowing down processing while waiting for notifications to be
processed and sent.
ok @davecgh
This mirrors a recent change in the reference implementation. Since the
vast majority of the network still uses it and transactions which aren't
considered standard won't be relayed, the rules for standard transactions
need to be kept consistent.
Ideally this should be handled via floating fees, but we're rather limited
until we have miner support and a larger network share so the probability
of a transaction being relayed and confirmed is high.
Closes#100.
ok @owainga
This commit exports the VarIntSerializeSize function to provide callers
with an easy method to determine how many bytes it would take to serialize
the passed value as a variable length integer.
This commit exports a new variant of the existing internal
calcNextRequiredDifficulty function which calculates and returns the next
required difficulty for a block after the end of the current best chain
based on the difficulty retarget rules.
In order to support the exported function the internal one was slightly
modified to accept the block timestamp instead of an entire block since
the timestamp is all that is needed and the caller of the exported
function might next have a full block yet.
This commit modifies the BuildMerkleTreeStore function to accept a slice
of btcutil.Tx transactions as opposed to a full block. This allows more
flexibility when calculating merkle roots since a full block may not be
created yet (particularly when generating blocks that need to be solved in
mining).
Previously, the BuildMerkleTreeStore function accepted a btcutil.Block
because originally the block itself cached the transaction hashes and it
was necessary to have access to the block to make use of the cached
transactions. However, the code has since been improved such that it
caches transaction hashes directly in each btcutil.Tx. This means the
code can remain as efficient as before while allowing the individual
transacitons to be passed.
This commit simply alphabetizes the subsystem logger variables, map, and
use switch so the order in the code is consistent with the sorted output
displayed when using --debuglevel.
The debuglevel parameter has accepted subsystems in additional to an
overall level for quite some time, but the sample config file was not
updated to reflect that.
This commit updates the sample config file accordingly.
This commit changes the websocket client code to use a mutex for
disconnect since it's theoretically possible a non-blocking select on the
quit channel could fall through from two different goroutines thus causing
a second call to close.
ok @jrick.
This change removes the processedtx notification, replacing it with
recvtx, and the spenttx notification, replacing it with redeemingtx.
Both new transactions return the full serialized transaction (encoded
in hexadecimal) rather than details about the transaction.
The old txmined notification has also been completely removed as it is
unreliable due to transaction malleability and no code should be
depending on it.
This change improves the mechanism by which btcd notifies a websocket
client of transaction relating to watched address and unspent outputs
in the following ways:
1. The old processedtx notification has been replaced with the new
recvtx notification. This notification, rather than parsing out
details used by wallet clients, sends the serialized transaction
(as hexadecimal) and any block details (if mined) if any transaction
output sends to one of the websocket client's watched addresses.
2. The old txspent notification has been replaced with the new
redeemingtx notification. This notification, rather than parsing
out details used by wallet clients, sends the serialized transaction
(as hexadecimal) and any block details (if mined) if any transaction
input spends a watched output.
3. When processing notifications for transaction outputs, if any output
spends to a client's watched address, a corresponding spent request
is automatically registered.
4. Transaction notifications originating from mempool now include both
transaction inputs and outputs, rather than only processing
5. When processing notifications for transaction inputs, a client's
output spent request is only removed if the transaction being
processed has also been mined into a block. In combination with the
4th change, this results in two redeemingtx notifications for
transactions which first appear in mempool and are subsequently mined
into a block.
This commit adds an additional sanity check to ensure the block that is
being processed does not contain a timestamp with a precision higher than
one second. This is necessary because the consensus rules only deal with
whole seconds in the time comparisons whereas the internal data structures
make use of Go time.Time values which support up to nanosecond precision.
Also, add a test to ensure the new functionality works as expected.
ok @owainga
This commit changes all cases which generate default timestamps to
time.Now to limit the timestamp to one second precision. The code which
serializes and deserializes timestamps already does this, but it is useful
to make sure defaults don't exceed the precision of the protocol either.
With this change there is less chance that developers using defaults will
end up with structures that have a higher time precision than what will
ultimately be sent across the wire.
This commit moves the subsidy halving interval to the chain params so it
can be configured per network. With that change it sets the regression
test halving interval to 150 to match the regression test params of the
reference implementation.
This was pointed out by @flammit.
The name findLatestKnownCheckpoint is confusing and doesn't really convey
the fact the purpose of the function is to the find the checkpoint prior
to the current known block.
Therefore, rename the function to findPreviousCheckpoint for clarity.
Also, update some comments to follow suit.
This commit adds an additional check to the block acceptance rules which
prevents new blocks that fork the main chain before the previous known
good checkpoint. This prevents storage of new, otherwise valid, blocks
from building off of old blocks which are likely at a much easier
difficulty and therefore could be used to waste cache and disk space.
Note this is slightly different than the other existing check which
prevents blocks with a timestamp before the timestamp of the latest known
good checkpoint since that check effectively prevents old side chain
blocks that already existed (per the claimed timestamp).
ok drahn@
This commit tightens the check for a pay-to-pubkey script by ensuring the
length of the pubkey is one of the two valid values of 33 or 65. This
mirrors the checks in the multisig script type check as well.
ok @owainga
This changes the implementation of the sendrawtransaction RPC handler
to match bitcoind behavior by always returning a rejection error for
any error processing or accepting the tx by the mempool. Previously,
if the tx was rejected for a rule error rather than an actual failure,
a client would still receive the tx sha as a result with no error.
Also, unexport the functions to generate script types. Everything should
(and is) be using PayToAddrScript() with an address type instead of
throwing bytes around.
discussed with #@davecgh
