691d45fdc8 Add coinstatsindex_unclean_shutdown test (Ryan Ofsky)
eb6cc05da3 index: Commit DB_MUHASH and DB_BEST_BLOCK to disk together (Martin Zumsande)
Pull request description:
Fixes #24076
Coinstatsindex currently writes the MuHash (`DB_MUHASH`) to disk in `CoinStatsIndex::WriteBlock()` and `CoinStatsIndex::ReverseBlock()`, but the best synced block is written in `BaseIndex::Commit()`. These are called at different points in time, both during the ThreadSync phase, and also after the initial sync is finished and validation callbacks (`BlockConnected()` vs `ChainStateFlushed()`) perform the syncing.
As a result, the index DB is temporarily in an inconsistent state, and if bitcoind is terminated uncleanly (so that there is no time to call `Commit()` by receiving an interrupt or by flushing the chainstate) this leads to problems:
On the next startup, `Init()` will read the best block and a MuHash that corresponds to a different (higher) block. Indexing will be picked up at the the best block processing some blocks again, but since MuHash is a rolling hash, it will process some utxos twice and the muhashes for all future blocks will be wrong, as was observed in #24076.
Fix this by always committing `DB_MUHASH` together with `DB_BEST_BLOCK`.
Note that the block data for the index is still written at different times, but this does not corrupt the index - at worst, these entries will be processed another time and overwritten after an unclean shutdown and restart.
ACKs for top commit:
ryanofsky:
Code review ACK 691d45fdc8. Only change since last review is adding test
fjahr:
ACK 691d45fdc8
Tree-SHA512: e1c3b5f06fa4baacd1b070abb0f8111fe2ea4a001ca8b8bf892e96597cf8b5d5ea10fa8fb837cfbf46648f052c742d912add4ce26d4406294fc5fc20809a0e1b
Fix GUI startup crash reported by Rspigler in
https://github.com/bitcoin/bitcoin/issues/24457 that happens if
settings.json contains an integer value for any of the configuration
options which GUI settings can currently clash with (-dbcache, -par,
-spendzeroconfchange, -signer, -upnp, -natpmp, -listen, -server, -proxy,
-proxy, -onion, -onion, -lang, and -prune).
Fix is a one-line change in ArgsManager::GetArg.
Just add tests. No changes to application behavior. Tests will be
updated in the next commit changing & improving current behavior.
Include a Qt test for GUI startup crash reported by Rspigler in
https://github.com/bitcoin/bitcoin/issues/24457 caused by GetArg
behavior that happens if settings.json contains an integer value for any
of the configuration options which GUI settings can currently clash with
(-dbcache, -par, -spendzeroconfchange, -signer, -upnp, -natpmp, -listen,
-server, -proxy, -proxy, -onion, -onion, -lang, and -prune).
6c23c41561 refactor: Rewrite AddToBlockIndex with try_emplace (Carl Dong)
c05cf7aa1e style: Modernize range-based loops over m_block_index (Carl Dong)
c2a1655799 style-only: Use using instead of typedef for BlockMap (Carl Dong)
dd79dad175 refactor: Rewrite InsertBlockIndex with try_emplace (Carl Dong)
531dce0347 tests: Remove now-unnecessary manual Unload's (Carl Dong)
bec86ae326 blockstorage: Make m_block_index own CBlockIndex's (Carl Dong)
Pull request description:
Part of: #24303
Split off from: #22564
```
Instead of having CBlockIndex's live on the heap, which requires manual
memory management, have them be owned by m_block_index. This means that
they will live and die with BlockManager.
```
The second commit demonstrates how this makes calls to `Unload()` to satisfy the address sanitizer unnecessary.
ACKs for top commit:
ajtowns:
ACK 6c23c41561
MarcoFalke:
re-ACK 6c23c41561🎨
Tree-SHA512: 81b2b5119be27cc0f8a9457b11da60cc60930315d2a5be36be89fe253d32073ffe622348ff153114b9b3212197bddbc791810913a43811b33cc58e7162bd105b
ae9ceed3e2 validation, refactoring: remove ChainstateManager::Reset() (Jon Atack)
daad0093e3 validation: replace lock with annotation in UnloadBlockIndex() (Jon Atack)
Pull request description:
Thread safety refactoring seen in #24177:
- replace re-acquiring lock cs_main with a thread safety annotation in UnloadBlockIndex()
- remove ChainstateManager::Reset(), as it is currently unused (can be reintroduced in the test utilities if needed for unit testing)
ACKs for top commit:
laanwj:
Code review ACK ae9ceed3e2
vasild:
ACK ae9ceed3e2
klementtan:
crACK ae9ceed3e2
Tree-SHA512: cebb782572997cc2dda01590d6bb6c5e479e8202324d8b6ff459b814ce09e818b996c881736bfebd1b8bf4b6d7a0f79faf3ffea176a4699dd7d7429de2db2d13
60aa179d8f Use GetPathArg where possible (Pavol Rusnak)
5b946edd73 util, refactor: Use GetPathArg to read "-settings" value (Ryan Ofsky)
687e655ae2 util: Add GetPathArg default path argument (Ryan Ofsky)
Pull request description:
Improve `ArgsManager::GetPathArg` method added in recent PR #24265, so it is usable more places. This PR starts to use it for the `-settings` option. This can also be helpful for #24274 which is parsing more path options.
- Add `GetPathArg` default argument so it is less awkward to use to parse options that have default values.
- Fix `GetPathArg` negated argument handling. Return path{} not path{"0"} when path argument is negated.
- Add unit tests for default and negated cases
- Move `GetPathArg` method declaration next to `GetArg` declaration. The two methods are close substitutes for each, so this should help keep them consistent and make them more discoverable.
ACKs for top commit:
w0xlt:
Tested ACK 60aa179 on Ubuntu 21.10
hebasto:
re-ACK 60aa179d8f
Tree-SHA512: 3d24b885d8bbeef39ea5d0556e2f09b9e5f4a21179cef11cbbbc1b84da29c8fb66ba698889054ce28d80bc25926687654c8532ed46054bf5b2dd1837866bd1cd
bbbbeaf9c8 fuzz: Limit script_format to 100kB (MarcoFalke)
Pull request description:
The target is still one of the slowest ones, but doesn't seem incredibly important. Especially for sizes larger than the standard tx size.
Fix that by limiting the script size.
ACKs for top commit:
fanquake:
ACK bbbbeaf9c8
Tree-SHA512: b6cf7248753909ef2f21d8824f187e7c05732dd3b99619c0067f862f3c2b0f9a87779d4ddbbd3a7a4bae5c794280e2f0a223bf835d6bc6ccaba01817d69479a2
Call fs::u8path to convert some UTF-8 string literals to paths, instead
of relying on implicit conversions. The implicit conversions incorrectly
decode const char* paths using the current windows codepage, instead of
treating them as UTF-8. This could cause test failures depending what
environment windows tests are run in.
Issue was reported by MarcoFalke <falke.marco@gmail.com> in
https://github.com/bitcoin/bitcoin/pull/24306#discussion_r818566106
`GetListenPort()` uses a simple logic: "if `-port=P` is given, then we
must be listening on `P`, otherwise we must be listening on `8333`".
This is however not true if `-bind=` has been provided with `:port` part
or if `-whitebind=` has been provided. Thus, extend `GetListenPort()` to
return the port from `-bind=` or `-whitebind=`, if any.
Fixes https://github.com/bitcoin/bitcoin/issues/20184 (cases 1. 2. 3. 5.)
Add a new function `TestOnlyResetTimeData()` which would reset the
internal state used by `GetTimeOffset()`, `GetAdjustedTime()` and
`AddTimeData()`.
This is needed so that unit tests that call `AddTimeData()` can restore
the state in order not to confuse other tests that rely on it.
Currently `timedata_tests/addtimedata` is the only test that modifies
the state (via `AddTimeData()`) and also the only test that relies on
that state.
5d7f22595f Do not use `LocalTestingSetup` in getarg_tests test file. (Kiminuo)
Pull request description:
Avoid using a test fixture in getarg_tests for better readability. Change was implemented by _kiminuo_ and posted https://github.com/bitcoin/bitcoin/pull/24306#issuecomment-1036643216
ACKs for top commit:
kiminuo:
ACK 5d7f22595f
Tree-SHA512: 0fd98622010e6923e91c66447a1d0861bf344a65d86a313dff7d428c089b1740a25f699327f6ed4c163255f270bcbd4f7be962bb551862214f9b9e395d40df04
Let GetPathArg method be used more places for path arguments that have
default values, like "-settings" and BITCOIN_SETTINGS_FILENAME in the
next commit.
Also:
- Fix negated argument handling. Return path{} not path{"0"} when path
argument is negated.
- Add new tests for default and negated cases
- Move GetPathArg() method declaration next to GetArg() declarations.
The two methods are close substitutes for each other, so this should
help keep them consistent and make them more discoverable.
b7be28cac5 test: add combined CJDNS/I2P/localhost/onion eviction protection tests (Jon Atack)
0a1bb84770 test: add tests for inbound eviction protection of CJDNS peers (Jon Atack)
0c00c0c981 test: fix off-by-one logic in an eviction protection test (Jon Atack)
f7b8094d61 p2p: extend inbound eviction protection by network to CJDNS peers (Jon Atack)
Pull request description:
Extend inbound eviction protection for peers connected over CJDNS, as is the case for peers connected via onion, localhost, and I2P since #21261 and #20197. CJDNS peers seem to have better min ping latency than onion and I2P peers but still higher than that of unencrypted IPv4/6 peers and can be disadvantaged under our eviction criteria. They are also very few in number, which is a further reason to protect them, as the goal of this logic is to favorise the diversity of our peer connections. CJDNS support was added in #23077 for the upcoming v23 release.
ACKs for top commit:
laanwj:
Concept and code review ACK b7be28cac5
w0xlt:
tACK b7be28c
Tree-SHA512: 89ebdd217602e16ae14b9bd0d5a25fc09f9b2384c951f820bc0f5a6d8452bbc9042065db817d5d5296c0ad22988491a83fc5b9a611e660c40ebd4f03448c4061
36ee76d1af net: remove unused CNetAddr::GetHash() (Vasil Dimov)
d0abce9a50 net: include the port when deciding a relay destination (Vasil Dimov)
2e38a0e686 net: add CServiceHash constructor so the caller can provide the salts (Vasil Dimov)
97208634b9 net: open p2p connections to nodes that listen on non-default ports (Vasil Dimov)
Pull request description:
By default, for mainnet, the p2p listening port is 8333. Bitcoin Core
has a strong preference for only connecting to nodes that listen on that
port.
Remove that preference because connections over clearnet that involve
port 8333 make it easy to detect, analyze, block or divert Bitcoin p2p
traffic before the connection is even established (at TCP SYN time).
For further justification see the OP of:
https://github.com/bitcoin/bitcoin/pull/23306
ACKs for top commit:
laanwj:
Concept and light code review ACK 36ee76d1af
prayank23:
ACK 36ee76d1af
stickies-v:
tACK 36ee76d1a
jonatack:
ACK 36ee76d1af
glozow:
utACK 36ee76d1af
Tree-SHA512: 7f45ab7567c51c19fc50fabbaf84f0cc8883a8eef84272b76435c014c31d89144271d70dd387212cc1114213165d76b4d20a5ddb8dbc958fe7e74e6ddbd56d11
fad7ddf9e3 test: Run symlink regression tests on Windows (MarcoFalke)
Pull request description:
Seems odd to add tests, but not run them on the platform that needs them most.
ACKs for top commit:
laanwj:
Code review ACK fad7ddf9e3
ryanofsky:
Code review ACK fad7ddf9e3, just removing new test. Would be nice if the test could be added later, of course.
Tree-SHA512: 64b235967a38c2eb90657e8d7a0447bcc8ce81d1b75a275b6c48bd42efd9ea7e7939257e484f297ee84598def3738eaeb289561aeba1dd6a99b258d389995139
fafc4eb363 test: Fix Wambiguous-reversed-operator compiler warnings (MarcoFalke)
Pull request description:
Add a missing const to avoid the C++20 clang **compiler warning**:
```
test/fuzz/addrman.cpp:325:22: error: ISO C++20 considers use of overloaded operator '==' (with operand types 'AddrManDeterministic' and 'AddrManDeterministic') to be ambiguous despite there being a unique best viable function [-Werror,-Wambiguous-reversed-operator]
assert(addr_man1 == addr_man2);
~~~~~~~~~ ^ ~~~~~~~~~
/usr/include/assert.h:93:27: note: expanded from macro 'assert'
(static_cast <bool> (expr) \
^~~~
test/fuzz/addrman.cpp:140:10: note: ambiguity is between a regular call to this operator and a call with the argument order reversed
bool operator==(const AddrManDeterministic& other)
^
1 error generated.
```
This patch also fixes the **compile error** if the first operand is `const`:
```
test/fuzz/addrman.cpp:326:23: error: invalid operands to binary expression ('const AddrManDeterministic' and 'AddrManDeterministic')
assert(addr_man_1 == addr_man2);
~~~~~~~~~~ ^ ~~~~~~~~~
/usr/include/assert.h:90:27: note: expanded from macro 'assert'
(static_cast <bool> (expr) \
^~~~
test/fuzz/addrman.cpp:140:10: note: candidate function not viable: 'this' argument has type 'const AddrManDeterministic', but method is not marked const
bool operator==(const AddrManDeterministic& other)
^
1 error generated.
ACKs for top commit:
hebasto:
ACK fafc4eb363, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: 92cd62ae06ee1393a6dc2ea6f3f553595a8f8d66f51592d231b42122bfb71ed4801a016daafc85360040339c5ae59b76888265cec37449c4688d6c7768f4567e
48742693ac Replace "can not" with "cannot" in docs, user messages, and tests (Jon Atack)
e670edd434 User-facing content fixups from transifex translator feedback (Jon Atack)
Pull request description:
Closes #24366.
ACKs for top commit:
laanwj:
Code review re-ACK 48742693ac
hebasto:
re-ACK 48742693ac, only suggested change since my previous [review](https://github.com/bitcoin/bitcoin/pull/24367#pullrequestreview-885938219).
Tree-SHA512: 4dcdcb417251a413e65fab6070515e13a1267c8e0dbcf521386b842511391f24c84a0c2168fe13458c977682034466509bf2a3453719d4d94d3c568fd9f4adb4
0683f377e1 Add tr() descriptor unit tests (Pieter Wuille)
4b2e31a7ae Bugfix: make ToPrivateString work with x-only keys (Pieter Wuille)
18ad54c3b2 Bugfix: set x-only flag when inferring pk() inside tr() (Pieter Wuille)
Pull request description:
This fixes two bugs in the current logic for `tr()` descriptors:
* ToPrivateString does not always work, because the provided private key may mismatch the parity of the x-only public key.
* The descriptors inferred for `pk()` inside `tr()` have the wrong x-only flag, leading to such descriptors generating the wrong scriptPubKey (roundtripping through ToString does fix it however, so this seems unobservable in the current code).
These were discovered while adding unit tests to descriptor_tests that cover various aspects of `tr()` descriptors, which are now also added here.
ACKs for top commit:
achow101:
ACK 0683f377e1
instagibbs:
ACK 0683f377e1
jonatack:
Code review ACK 0683f377e1
Tree-SHA512: fc0e11b45da53054a108effff2029d67b64e508b160a6e22e00c98b506c39ec12ccc95afd21ea68a6c691eb62930afc7af18908f2fa3a954d102afdc67bc355a
fae3f17823 fuzz: Split script formatting from script fuzz target (MarcoFalke)
Pull request description:
This is a follow-up to commit 9237bdaac1.
The target was improved a bit, but is still taking enormously long. See for example 4096 seconds in https://cirrus-ci.com/task/5153886888525824?logs=ci#L4451.
Most of the time is spent formatting the script. See the flamegraph: 
Thus, I suggest to split up the formatting into a new target. This will:
* Allow more fuzz cycles in the `script` target when exploring the search space with the fuzz engine
* Hopefully allow to reduce the fuzz inputs in `qa-assets` without losing coverage
ACKs for top commit:
fanquake:
ACK fae3f17823
Tree-SHA512: f86154b23019b7721e5dd10f54d11f4f7603d280471a396cb5256f4c460f48333318a60efe8b77fa8749a4abc67ad2631211b766fde5da70ded9fab8f904747b
Work around libstdc++ issue [PR101510] with create_directories where the
leaf already exists as a symlink. Fixes #24257, introduced by the switch
to `std::filesystem`. It is meant to be more thorough than #24266, which
only worked around one instance of the problem.
The issue was fixed upstream in
https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=124eaa50e0a34f5f89572c1aa812c50979da58fc,
but unfortunately we'll have to carry a fix for it for a while.
This introduces a function `fs::create_directories` which wraps
`std::filesystem::create_directories`. This allows easiliy reverting the
workaround when it is no longer necessary.
dc01cbc538 test: Add fs_tests/rename unit test (Hennadii Stepanov)
d4999d40b9 util: Revert back MoveFileExW call for MinGW-w64 (Hennadii Stepanov)
Pull request description:
Unfortunately, bitcoin/bitcoin#24308 introduced a [regression](https://github.com/bitcoin/bitcoin/pull/24308#issuecomment-1037259386) for mingw builds.
The root of the problem is a broken implementation of [`std::filesystem::rename`](https://en.cppreference.com/w/cpp/filesystem/rename). In particular, the expected behavior
> If `old_p` is a non-directory file, then `new_p` must be ... existing non-directory file: `new_p` _is first deleted_...
fails with the "File exists" error.
This PR reverts back the `MoveFileExW` call, and adds the [suggested](https://github.com/bitcoin/bitcoin/pull/24308#pullrequestreview-878832906) unit test.
ACKs for top commit:
vasild:
ACK dc01cbc538
Tree-SHA512: c8e5a98844cfa32bec0ad67a1aaa58fe2efd0c5474d3e83490211985b110f83245758a742dcaa0a933a192ab66a7f11807e0c53ae69260b7dd02fc99f6d03849
34d0e07e92 Test that OP_1-OP_16 (but not lower/higher) start witness programs (Pieter Wuille)
Pull request description:
Cherry-picks one of the commits adding test coverage from #13062. As [pointed out by aj](https://github.com/bitcoin/bitcoin/pull/13062/files#r492723037):
> could move the test additions to the first commit, since they're testing things that are already true
Pull the additional test code into master earlier.
ACKs for top commit:
laanwj:
Code review ACK 34d0e07e92
Tree-SHA512: ff0ab2a54613ea6e8246b443363b362dd41b5e464faba4d11be6003aa6588a626cf56e142a3b94465cd37dd3ac4debea08455db96bade336171b6c30ea894950
This new constructor will be useful if we just want to hash a `CService`
object without the two `GetRand()` calls (in `RelayAddress()` in a
subsequent commit).
By default, for mainnet, the p2p listening port is 8333. Bitcoin Core
has a strong preference for only connecting to nodes that listen on that
port.
Remove that preference because connections over clearnet that involve
port 8333 make it easy to detect, analyze, block or divert Bitcoin p2p
traffic before the connection is even established (at TCP SYN time).
For further justification see the OP of:
https://github.com/bitcoin/bitcoin/pull/23306
fa2f7d0059 fuzz: Avoid unsigned integer overflow in FormatParagraph (MarcoFalke)
Pull request description:
`FormatParagraph` is only ever called with compile time constant arguments, so I don't see the need for fuzzing it.
Though, keep it for now, but avoid the unsigned integer overflow with this patch.
ACKs for top commit:
laanwj:
Code review ACK fa2f7d0059
Tree-SHA512: 01fc64a9ef73c183921ca1b0cd8db9514c0a242e3acf215a3393f383ae129e01625ebb16eaf9cb86370eda62d0145c3dcf8f62e40edf5958abc1f777c5687280
fa1b227a72 Remove broken and unused CDataStream methods (MarcoFalke)
faee5f8dc2 test: Create fresh CDataStream each time (MarcoFalke)
fa71114926 test: Inline expected_xor (MarcoFalke)
Pull request description:
The `insert` and `erase` methods have many issues:
* They are unused
* They are confusing and hard to read, as they implement "special cases" for optimization, that isn't needed
* They are broken (See https://github.com/bitcoin/bitcoin/pull/24231)
* Fixing them leads to mingw compile errors (See https://github.com/bitcoin/bitcoin/pull/24231#issuecomment-1029286985)
Fix all issues by removing them
ACKs for top commit:
laanwj:
Code review ACK fa1b227a72
Tree-SHA512: 9d9e5d42e6ffc5ae82bdb67cfb5b50b45977ae674acee6ff99092560aebf2fc7e4584ded614e190db0663226fa198e34350517cd7ee57d518de22e9568bc349a
