a59dac3 refactor: Avoid extra lookups of mapAddressBook in listunspent RPC (João Barbosa)
d76962e rpc: Reduce cs_main lock in listunspent (João Barbosa)
Pull request description:
On my system, where the wallet has 10000 unspents, the `cs_main` lock duration changed from 191ms to 36ms. The loop that generates the response takes around 155ms. So, the lock duration is reduced to around 20%.
Tree-SHA512: ddaae591f39da59a9d1a8e9ffe773d857687789476f566ca273d310ad531da6dacff80cac69f3334c601c251ac7c5ed4136656c725aa3d611c6bbf734111946e
1e46d8a Get rid of ambiguous OutputType::NONE value (Russell Yanofsky)
Pull request description:
Based on suggestion by @sipa https://github.com/bitcoin/bitcoin/pull/12119#issuecomment-357982763
After #12119, the NONE output type was overloaded to refer to either an output type that couldn't be parsed, or to an automatic change output mode. This change drops the NONE enum and uses a simple bool to indicate parse failure, and a new CHANGE_AUTO enum to refer the change output type.
This change is almost a pure refactoring except it makes RPCs reject empty string ("") address types instead of treating them like they were unset. This simplifies the parsing code a little bit and could prevent RPC usage mistakes. It's noted in the release notes.
Follows up #12408 by @MarcoFalke
Followups for future PRs:
- [ ] Add explicit support for specifying "auto" in `ParseOutputType` as suggested by promag and sipa: https://github.com/bitcoin/bitcoin/pull/12729#issuecomment-374799567 and https://github.com/bitcoin/bitcoin/pull/12729#discussion_r175969481
- [ ] Add wallet `AddressChangeType` method to complement `TransactionChangeType`: https://github.com/bitcoin/bitcoin/pull/12729#discussion_r175969618.
Tree-SHA512: 8b08b272bcb177a0a9e556dcd965840a7fe601ef83ca97938b879c9b1a33b5b3f96939e1bceef11ba7c644ac21bfd6c1dbc6ca715cd1da4ace50475240e4ee48
3fdc5fe Make sure initialization occurs in the constructor (practicalswift)
1e7813e Remove redundant initializations from the constructor (practicalswift)
f131872 Initialize non-static class members where they are defined (practicalswift)
73bc1b7 Initialize editStatus and autoCompleter. Previously not initialized where defined or in constructor. (practicalswift)
Pull request description:
Initialize variables previously neither defined where defined nor in constructor:
* `editStatus`
* `autoCompleter`
Also; initialize non-static class members where they are defined in accordance with developer notes.
Tree-SHA512: 84f0cb87ec8394ed7641bfa0731be2ec72e6a920e00ae206ff89e2e7c960358f603c52878311b24601a33aa7cba6ea4f9a78a8ade88112dea0f41efb08e84e25
9cb6cdc Simplify semantics of ChainStateFlushed callback (Matt Corallo)
50b6533 scripted-diff: Rename SetBestChain callback ChainStateFlushed (Matt Corallo)
Pull request description:
Previously, ChainStateFlushed would fire either if a full flush
completed (which can happen due to memory limits, forced flush, or
on its own DATABASE_WRITE_INTERVAL timer) *or* on a
ChainStateFlushed-specific DATABASE_WRITE_INTERVAL timer. This is
both less clear for clients (as there are no guarantees about a
flush having actually happened prior to the call), and reults in
extra flushes not clearly intended by the code. We drop the second
case, providing a strong guarantee without removing the periodit
timer-based flushing.
This is a follow-up to discussion in #11857.
Tree-SHA512: 22ba3a0954d265d28413dbf87040790ca5b439820ee7bbadab14028295ec190de82ce5cd664426c82e58b706dc84278868026fa8d066702eb6e6962c9ace1f8e
`LogPrintStr` returns the number of characters printed. This number is
doubled if both logging to console and logging to file is enabled. As
the return value is never used, I've opted to remove it instead of try
to fix it.
Credit: @laanwj
8c2d695c4a util: Store debug log file path in BCLog::Logger member. (Jim Posen)
8e7b961388 scripted-diff: Rename BCLog::Logger member variables. (Jim Posen)
1eac317f25 util: Refactor GetLogCategory. (Jim Posen)
3316a9ebb6 util: Encapsulate logCategories within BCLog::Logger. (Jim Posen)
6a6d764ca5 util: Move debug file management functions into Logger. (Jim Posen)
f55f4fcf05 util: Establish global logger object. (Jim Posen)
Pull request description:
This is purely a refactor with no behavior changes.
This creates a new class `BCLog::Logger` to encapsulate all global logging configuration and state.
Tree-SHA512: b34811f54a53b7375d7b6f84925453c6f2419d21179379ee28b3843d0f4ff8e22020de84a5e783453ea927e9074e32de8ecd05a6fa50d7bb05502001aaed8e53
e2f58f4 wallet: Make vpwallets usage thread safe (João Barbosa)
Pull request description:
This PR turns the functions introduced in #13017 thread safe. This is required to correctly support dynamically loading wallets, which is implemented in #10740.
Tree-SHA512: efaa09e501636cf957aa33de83719ce09dc0c2a19daff741a94ef10d6b7ba5dee538355b80c96ead995140f99f5df0c92fb0e22ae1adb8f397eb478280c8d8c7
66dc662c8a Add compile time checking for all cs_KeyStore runtime locking assertions (practicalswift)
Pull request description:
Add compile time checking for all `cs_KeyStore` runtime locking assertions.
This PR is a subset of #12665. The PR was broken up to make reviewing easier.
The intention is that literally all `EXCLUSIVE_LOCKS_REQUIRED`/`LOCKS_EXCLUDED`:s added in this PR should follow either directly or indirectly from `AssertLockHeld(…)`/`AssertLockNotHeld(…)`:s already existing in the repo.
Consider the case where function `A(…)` contains `AssertLockHeld(cs_foo)` (without
first locking `cs_foo` in `A`), and that `B(…)` calls `A(…)` (without first locking `cs_main`):
* It _directly_ follows that: `A(…)` should have an `EXCLUSIVE_LOCKS_REQUIRED(cs_foo)` annotation.
* It _indirectly_ follows that: `B(…)` should have an `EXCLUSIVE_LOCKS_REQUIRED(cs_foo)` annotation.
Tree-SHA512: 84ee5459e7f75f9affaa4275cb760576ab0e26da8a48b9a4a59b51a25418fe4b862ba832cb01312479a8c9b0e38ee8b6c4076bcbbd73213346d09ec2057fc9f1
Previously, ChainStateFlushed would fire either if a full flush
completed (which can happen due to memory limits, forced flush, or
on its own DATABASE_WRITE_INTERVAL timer) *or* on a
ChainStateFlushed-specific DATABASE_WRITE_INTERVAL timer. This is
both less clear for clients (as there are no guarantees about a
flush having actually happened prior to the call), and reults in
extra flushes not clearly intended by the code. We drop the second
case, providing a strong guarantee without removing the periodit
timer-based flushing.
This much more accurately captures the meaning of the callback.
-BEGIN VERIFY SCRIPT-
sed -i 's/SetBestChain/ChainStateFlushed/g' src/validationinterface.h src/validationinterface.cpp src/wallet/wallet.h src/wallet/wallet.cpp src/validation.cpp src/index/txindex.h src/index/txindex.cpp
-END VERIFY SCRIPT-
d8e9a2a Remove "rpc" category from GetWarnings (Wladimir J. van der Laan)
7da3b0a rpc: Move RPC_FORBIDDEN_BY_SAFE_MODE code to reserved section (Wladimir J. van der Laan)
2ae705d Remove Safe mode (Andrew Chow)
Pull request description:
Rebase of #10563. Safe mode was [disabled by default and deprecated in 0.16](https://github.com/bitcoin/bitcoin/blob/master/doc/release-notes/release-notes-0.16.0.md#safe-mode-disabled-by-default), so probably should be removed for 0.17.
> Rationale:
>
> Safe mode is useless. It only disables some RPC commands when large work forks are detected. Nothing else is affected by safe mode. It seems that very few people would be affected by safe mode. The people who use Core as a wallet are primarily using it through the GUI, which safe mode does not effect. In the GUI, transactions will still be made as normal; only a warning is displayed.
>
> I also don't think that we should be disabling RPC commands or any functionality in general. If we do, it should be done consistently, which safe mode is not. If we want to keep the idea of a safe mode around, I think that the current system needs to go first before a new system can be implemented.
Tree-SHA512: 067938f47ca6e879fb6c3c4e21f9946fd7c5da3cde67ef436f1666798c78d049225b9111dc97064f42b3bc549d3915229fa19ad5a634588f381e34fc65d64044
Although this code is no longer ever sent back after removing safe mode,
it would be unwise to remove it from the header.
For one, it would be bad to accidentally re-use the number.
Also some API documentation / bindings are directly generated from the .h
file - this is why the "Aliases for backward compatibility" are there. We don't
want to break code that relies on this error code existing, even if it's never
generated.
So keep it around but move it to a reserved section.
7d0f80b Use anonymous namespace instead of static functions (Pieter Wuille)
b61fb71 Mention removal of bare multisig IsMine in release notes (Pieter Wuille)
9c2a8b8 Do not treat bare multisig as IsMine (Pieter Wuille)
08f3228 Optimization: only test for witness scripts at top level (Pieter Wuille)
3619735 Track difference between scriptPubKey and P2SH execution in IsMine (Pieter Wuille)
ac6ec62 Switch to a private version of SigVersion inside IsMine (Pieter Wuille)
19fc973 Do not expose SigVersion argument to IsMine (Pieter Wuille)
fb1dfbb Remove unused IsMine overload (Pieter Wuille)
952d821 Make CScript -> CScriptID conversion explicit (Pieter Wuille)
Pull request description:
Currently our wallet code will treat bare multisig outputs (meaning scriptPubKeys with multiple public keys + `OP_CHECKMULTISIG` operator in it) as ours without the user asking for it, as long as all private keys in it are in our wallet.
This is a pointless feature. As it only works when all private keys are in one place, it's useless compared to single key outputs (P2PK, P2PKH, P2WPKH, P2SH-P2WPKH), and worse in terms of space, cost, UTXO size, and ability to test (due to lack of address format for them).
Furthermore, they are problematic in that producing a list of all `scriptPubKeys` we accept is not tractable (it involves all combinations of all public keys that are ours). In further wallet changes I'd like to move to a model where all scriptPubKeys that are treated as ours are explicit, rather than defined by whatever keys we have. The current behavior of the wallet is very hard to model in such a design, so I'd like to get rid of it.
I think there are two options:
* Remove it entirely (do not ever accept bare multisig outputs as ours, unless watched)
* Only accept bare multisig outputs in situations where the P2SH version of that output would also be acceptable
This PR implements the first option. The second option was explored in #12874.
Tree-SHA512: 917ed45b3cac864cee53e27f9a3e900390c576277fbd6751b1250becea04d692b3b426fa09065a3399931013bd579c4f3dbeeb29d51d19ed0c64da75d430ad9a
abd58a2 Fix for utiltime to compile with msvc. (Aaron Clauson)
Pull request description:
This PR allows utiltime.cpp to compile with msvc after the changes introduced in #12973.
Tree-SHA512: 7233b1c23400bf19aef2fcb6168009ef58b9e7f8e49c46d8cf9d04394091f370e39496d24ca00b294c4164bcfc04514e329bf6bb05169406c34ce7cd8c382565
41ff967 list the types of scripts we should consider for a witness program (fivepiece)
4f933b3 p2wpkh, p2wsh and p2sh-nested scripts in decodescript (fivepiece)
Pull request description:
Attempts to address #12244 . `p2wsh` addresses are returned only for scripts that are neither `p2sh` nor any witness program.
Tree-SHA512: eb47f094c1a4c2ad2bcf27a8032307e43cf787d50bf739281aeb4101d97316a2f307b05118bf138298c937fa34e15f91436443a9b313f809fad2c43e94cd1831
7de1de7 Add new fee structure with all sub-fields denominated in BTC (mryandao)
Pull request description:
the denomination for `fee` is current in btc while the other such as `decendentFee` and `ancestorFee` are in satoshis.
Tree-SHA512: e428f6dca1d339f89ab73e38ce5903f5465c46b159069d9bcc3f8b1140fe6657fa49a11abe0088e9f7ba9999f64af72a349a4735bf5eaa61b8e4a185b23543f3
Now that the transaction index is updated asynchronously, in order to
preserve the current behavior of public interfaces, the code blocks
until the transaction index is caught up with the current state of the
blockchain.
In order to preserve getrawtransaction RPC behavior, there needs to be
a way for a thread to ensure the transaction index is in sync with the
current state of the blockchain.
The TxIndex will be responsible for building the transaction index
concurrently with the main validation thread by implementing
ValidationInterface. This does not process blocks concurrently yet.
