In this commit, we add an additional test case for inherited RBF
replacement. This test case asserts that if a parent is marked as being
replaceable, but the child isn't, then the child can still be replaced
as according to BIP 125 it shoudl _inhreit_ the replaceability of its
parent.
The addition of this test case was prompted by the recently discovered
Bitcoin Core "CVE" [1]. It turns out that bitcoind doesn't properly
implement BIP 125. Namely it fails to allow a child to "inherit"
replaceability if its parent is also replaceable. Our implementation
makes this trait rather explicit due to its recursive implementation.
Kudos to the original implementer @wpaulino for getting this correct.
[1]: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2021-May/018893.html.
On signet all previous soft forks and also taproot are always activated,
meaning the version is always 0x20000000 for all blocks. To make sure
they activate properly in `btcd` we therefore need to use the correct
bit to mask the version.
This means that on any custom signet there would need to be 2016 blocks
mined before SegWit or Taproot can be used.
The previous use of allowSelfConns prevented this, as users aren't able
to invoke peer.TstAllowSelfConns themselves due to being part of a test
file, which aren't exported at the library level, leading to a
"disconnecting peer connected to self" error upon establishing a mock
connection between two peers. By including the option at the config
level instead (false by default, prevents connections to self) we enable
users of the peer library to properly test the behavior of the peer.Peer
struct externally.
When processRequest can't find a rpc command, standardCmdResult returns
a `btcjson.ErrRPCMethodNotFound` but it gets ignored and a
`btcjson.ErrRPCInvalidRequest` is returned instead.
This makes processRequest return the right error message.
When you provide an argument to EstimateFee(numblocks uint32) that exceeds the estimateFeeDepth (which is set to 25), you get an error message that says "can only estimate fees for up to 100 blocks from now". The variable used in the if condition and the variable used for creating the error message should be the same.
This resolves the more fundamental flake in the unit tests noted in the
prior commit.
Because multiple unit tests call rand.Seed in parallel, it's possible
they can be executed with the same unix timestamp (in seconds). If the
second call happens between generating the hash cache and checking that
the cache doesn't contain a random txn, the random transaction is in
fact a duplicate of one generated earlier since the RNG state was reset.
To remedy, we initialize rand.Seed once in the init function.
TestHashCacheAddContainsHashes flakes fairly regularly when rebasing
PR #1684 with:
txid <txid> wasn't inserted into cache but was found.
With probabilty 1/10^2 there will be no inputs on the transaction. This
reduces the entropy in the txid, and I belive is the primary cause of
the flake.
- create benchmarks to measure allocations
- add test for benchmark input
- create a low alloc parseScriptTemplate
- refactor parsing logic for a single opcode
Update the fields of GetBlockChainInfoResult to reflect the current state of
the RPC returned by other full-node implementations.
* InitialBlockDownload - Node is in Initial Block Download mode if True.
* SizeOnDisk - The estimated size of the block and undo files on disk.
The PR #1594 introduced a change that made the order of parameters
relevant, if one of them is nil. This makes it harder to be backward
compatible with the same JSON message if an existing parameter in
bitcoind was re-purposed to have a different meaning.
To allow using a custom btcd executable, we allow specifying a path to a
file. If the path is empty, the harness will fall back to compiling one
from scratch.
