fae20e6b50 Revert "Avoid the use of P0083R3 std::set::merge" (MarcoFalke)
fab53b5fd4 ci/doc: Set minimum required clang/libc++ version to 8.0 (MarcoFalke)
Pull request description:
This is not for 23.0, but for 24.0. It comes with the following benefits:
* Can use C++17 P0083R3 std::set::merge from libc++ 8.0
* No longer need to provide support for clang-7, which already fails to compile on some architectures (https://github.com/bitcoin/bitcoin/issues/21294#issuecomment-998098483)
This should be fine, given that all supported operating systems ship with at least clang-10:
* CentOS 8: clang-12
* Stretch: https://packages.debian.org/stretch/clang-11
* Buster: https://packages.debian.org/buster-backports/clang-11
* Bionic: https://packages.ubuntu.com/bionic-updates/clang-10
* Focal: https://packages.ubuntu.com/focal/clang-10
ACKs for top commit:
fanquake:
ACK fae20e6b50 - I think this is fine to do. I would be surprised if in another 6 months time someone was stuck on a system we supported, needing to compile Core, and only had access to Clang 7 or older. As mentioned in the PR description, all systems we currently support, already support multiple newer versions of Clang.
hebasto:
ACK fae20e6b50, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: 3b4c6c130ff40dd7e84934af076863415e5dd661d823c72e3e3832566c65be6e877a7ef9164bbcf394bcea4b897fc29a48db0f231c22ace0e2c9b5638659a628
ec7d73628a [wallet] assert BnB internally calculated waste is the same as GetSelectionWaste() (glozow)
Pull request description:
#22009 introduced a `GetSelectionWaste()` function to determine how much "waste" a coin selection solution has, and is a mirror of the waste calculated inside of `SelectCoinsBnB()`. It would be bad for these two waste metrics to deviate, since it could negatively affect how often we select the BnB solution. Add an assertion to help tests catch a potential accidental change.
ACKs for top commit:
achow101:
ACK ec7d73628a
Xekyo:
ACK ec7d73628a
Tree-SHA512: 3ab7ad45ae92e7ce3f21824fb975105b6be8382edf47c252df5d13d973a3abdcb675132d223b42fcbb669cca879672c904b8a58d0676e12bf381a9219f4db37c
e8272024ab doc: Use human-friendly DefaultHint for change_address/changeAddress in wallet RPC help (Luke Dashjr)
9d5e693c9d Bugfix: doc: Correct type of change_address/changeAddress in wallet RPC help (STR, not STR_HEX) (Luke Dashjr)
Pull request description:
ACKs for top commit:
achow101:
ACK e8272024ab
Tree-SHA512: da4db2b241160c93ea66f8c572c69d4688f52a5fd8c32b66b1192925fcb360baf91be9771eaca178f5b08e1920559174260ed57caddcffade48156ec0c83c0bc
40e871d9b4 [miner] always assume we can create witness blocks (glozow)
Pull request description:
Given the low possibility of a reorg reverting the segwit soft fork, there is no longer a need to check whether segwit is active to see if it's okay to add to the block template (see also #23512, #21009, etc). `TestBlockValidity()` is also run on the block template at the end of `CreateNewBlock()`, so any invalid block would be caught there.
ACKs for top commit:
gruve-p:
ACK 40e871d9b4
jnewbery:
utACK 40e871d9b4, although I disagree about changing the test for segwit transaction in mempool before activagtion, instead of just removing it: https://github.com/bitcoin/bitcoin/pull/24421#discussion_r822933721.
achow101:
ACK 40e871d9b4
theStack:
Code-review ACK 40e871d9b4
Tree-SHA512: bf4860bf2bed8339622d05228d11d60286edb0c32a9a3c434b8d154913c07ea56e50649f4af7009c2a1c6a58a81d2299ab43b41a6f16dee7d08cc89cc1603019
These two implementations of waste calculation should never deviate.
Still keep the SelectCoinsBnB internal calculation because incremental
calculate-as-you-go is much more performant than calling
GetSelectionWaste() over and over again.
8336a06dbd doc: remove Boost LDFLAGS from netBSD build docs (fanquake)
Pull request description:
We no-longer link against any Boost libs, so we shouldn't need to use
any Boost linker flags.
ACKs for top commit:
hebasto:
ACK 8336a06dbd, I have reviewed the code and it looks OK, I agree it can be merged. Also verified that there is no other usage of `BOOST_LDFLAGS` in our codebase or documentation.
Tree-SHA512: b7814d10cee789903cb3c613631e184a72f5766cda85261b5f99f9ac207348a2a49c92494c8c1d50163494f6b755c503cf51bf083b31f564dae1b0f493c54c2e
f1f994a122 doc: Add `guix` prefix for changes to reproducible builds (Hennadii Stepanov)
Pull request description:
Most of contributors already use the `guix:` prefix for changes to `contrib/guix`. Also `guix` is shorter than `build`, and it is more focused/specific.
ACKs for top commit:
fanquake:
ACK f1f994a122
Tree-SHA512: 3f754e80802ec4e871b099ce1f0877e34ecc4816fbe9c49bfd2a7368ef79fed9edf6c65f38eedef2a87367fdc911dc548e0def422d80b66a91ce2e5f35826032
1d4157a42b build: Fix Boost.Process detection on macOS arm64 (Hennadii Stepanov)
Pull request description:
Could be tested as follows:
```
% brew install boost@1.76
% ./autogen.sh
% ./configure --with-boost='/opt/homebrew/opt/boost@1.76'
```
ACKs for top commit:
promag:
Tested ACK on 1d4157a42b with boost 1.76 on macOS arm64. #24523 is required for boost 1.78.
Tree-SHA512: 7abd39a78e970ecc051e53b5923dfc31d3f0576cf4ff7fcfb9c8708c857c46a7a595ec36238def83f41158970eeee209980da4b8b70f0ff68f940a38ac9a0471
This was changed in #22818 from 0 to 1. Since it changes
BLOCK_OPT_WIT of the genesis block, older versions of bitcoin
core would not read regtest directories created with newer versions
without a reindex.
We don't use the deprecated headers now, and never should do in the
future, so there is no need for them to exist in depends.
The headers themselves are just full of includes for the newer headers.
b7dbc83f23 qt: Remove network detection based on address in BIP21 (laanwj)
Pull request description:
This is removes some ugly and brittle code that switches the global network to testnet based on a provided address. I think in practice it's very unlikely for testnet BIP21 payment URIs to be used, and if so it's for testing so it's easy enough to manually copy it. Or to specify `-testnet` explicitly.
There is already no such case for `-regtest` or `-signet`.
After this change it will only accept addresses for the explicitly selected network. Others will result in a "wrong network" popup.
There is also a possibility for refactor after this as the initialization order of `PaymentServer::ipcParseCommandLine` isn't important anymore (well, it still has to be before `PaymentServer::ipcSendCommandLine`, maybe even merged with it), but I have not done so here.
ACKs for top commit:
jonatack:
ACK b7dbc83f23
achow101:
ACK b7dbc83f23
Tree-SHA512: ebc77c0e5c98f53fe254bcd0f94c9d1c06937b794346e95b14158860d5c607515e71d73b782d2726674dce86d6d4001085d473c6d1bc11c5e6c25ff3fb3cfa22
3566353c5e ci: remove compiled-but-unused BDB from MSAN job (fanquake)
Pull request description:
Self-compiled BDB was added to this job as opposed to using depends BDB [due to linking issues](https://github.com/bitcoin/bitcoin/pull/18288#discussion_r433189350), however the compiled BDB is not actually used. Remove it for now, given we don't actually lose any coverage (note that BDB is also not currently used in the naitve MSAN fuzz job or for [OSS Fuzz](https://github.com/google/oss-fuzz/blob/master/projects/bitcoin-core/build.sh#L32) builds).
In future, we can use depends BDB, however introducing it now will cause false positives, which can be fixed by upgrading the versions of Clang / LLVM we use, however upgrading to those newer versions causes other issues, which appear in standard library code, and require more involved suppressing, which can be solved in a follow up or another PR i.e #23008.
Top commit has no ACKs.
Tree-SHA512: 9e8fdd95246cafa27cda7bcf0641b428d4573f6748ecdf07cc6205a64351db22ba383ec943e88a69df3694ccb9f125d994b64345a4e44fb6fea4a014504760d1
faa1aec26b Remove confusing P1008R1 violation in ATMPArgs (MarcoFalke)
Pull request description:
The `= delete` doesn't achieve the stated goal and it is also redundant, since it is not possible to default construct the `ATMPArgs` type.
This can be tested with:
```diff
diff --git a/src/validation.cpp b/src/validation.cpp
index 2813b62462..1c939c0b8a 100644
--- a/src/validation.cpp
+++ b/src/validation.cpp
@@ -519,6 +519,7 @@ public:
/** Parameters for child-with-unconfirmed-parents package validation. */
static ATMPArgs PackageChildWithParents(const CChainParams& chainparams, int64_t accept_time,
std::vector<COutPoint>& coins_to_uncache) {
+ ATMPArgs{};
return ATMPArgs{/* m_chainparams */ chainparams,
/* m_accept_time */ accept_time,
/* m_bypass_limits */ false,
```
Which fails on current master *and* this pull with the following error:
```
validation.cpp:525:22: error: reference member of type 'const CChainParams &' uninitialized
ATMPArgs{};
~^
validation.cpp:470:29: note: uninitialized reference member is here
const CChainParams& m_chainparams;
^
1 error generated.
```
Further reading (optional):
* http://open-std.org/JTC1/SC22/WG21/docs/papers/2018/p1008r1.pdf
ACKs for top commit:
achow101:
ACK faa1aec26b
glozow:
code review ACK faa1aec26b
Tree-SHA512: 16db2c9959a1996eafbfa533dc4d1483761b9d28295aed5a82b86abd7268da37c51c59ddc67c205165ecb415dbe637b12a0e1b3234d50ab0b3b79de66d7bd73e
Self-compiled BDB was added to this job as opposed to using depends BDB
due to linking issues, however the compiled BDB is not actually used.
Remove it for now, given we don't actually lose any coverage (note that
BDB is also no used the MSAN fuzz job), and in future, we can use
depends BDB.
8ea6167099 wallet: refactor: dedup sqlite blob binding (Sebastian Falbesoner)
Pull request description:
This refactoring PR deduplicates repeated SQLite blob binding to a statement with a newly introduced helper function `BindBlobToStatement`, abstracting away the calls to `sqlite3_bind_blob(...)`.
This should be more readable and less error-prone, e.g. in case that the error handling has to be adapted. As a slight drawback, the function where the binding happens is not printed anymore (`__func__`), i.e. one could argue this is not strictly a refactoring, but IMHO the advantages of deduplication outweigh this; binding errors are purely internal logic errors (wrong use of the sqlite API) rather than something that is dependend on external data like DB content.
ACKs for top commit:
laanwj:
Code review ACK 8ea6167099
achow101:
ACK 8ea6167099
klementtan:
ACK 8ea6167099
Tree-SHA512: 1de0d214f836bc405a01e98a3a2d71f2deaddc7d23c31cad80219d1614bec92619c06d9a4a091dd563d3e95ffb879649c29745d8f89669b2a5330552c212af3f
This is some very ugly and brittle code that switches the global network
based on a provided address, remove it. I think in practice it's very
unlikely for testnet BIP21 payment URIs to be used, and if so it's for
testing so it's easy enough to manually copy it. Or to specify
`-testnet` explicitly.
There is already no case for `-regtest` or `-signet`.
2f5fd3cf92 test: Correctly decode UTF-8 literal string paths (Ryan Ofsky)
Pull request description:
Call `fs::u8path()` to convert some UTF-8 string literals to paths, instead of relying on the implicit conversion. Fake Macro pointed out in https://github.com/bitcoin/bitcoin/pull/24306#discussion_r818566106 that `fs_tests` are incorrectly decoding some literal UTF-8 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 under.
The `fs::path` class exists to avoid problems like this, but because it is lenient with `const char*` conversions, under assumption that they are ["safe as long as the literals are ASCII"](727b0cb592/src/fs.h (L39)), bugs like this are still possible.
If we think this is a concern, followup options to try to prevent this bug in the future are:
0. Do nothing
1. Improve the "safe as long as the literals are ASCII" comment. Make it clear that non-ASCII strings are invalid.
2. Drop the implicit `const char*` conversion functions. This would be nice because it would simplifify the `fs::path` class a little, while making it safer. Drawback is that it would require some more verbosity from callers. For example, instead of `GetDataDirNet() / "mempool.dat"` they would have to write `GetDataDirNet() / fs::u8path("mempool.dat")`
3. Keep the implicit `const char*` conversion functions, but make them call `fs::u8path()` internally. Change the "safe as long as the literals are *ASCII*" comment to "safe as long as the literals are *UTF-8*".
I'd be happy with 0, 1, or 2. I'd be a little resistant to 3 even though it was would add more safety, because it would slightly increase complexity, and because I think it would encourage representing paths as strings, when I think there are so many footguns associated with paths as strings, that it's best to convert strings to paths at the earliest point possible, and convert paths to strings at the latest point possible.
ACKs for top commit:
laanwj:
Code review ACK 2f5fd3cf92
w0xlt:
crACK 2f5fd3c
Tree-SHA512: 9c56714744592094d873b79843b526d20f31ed05eff957d698368d66025764eae8bfd5305d5f7b6cc38803f0d85fa5552003e5c6cacf1e076ea6d313bcbc960c
e8023100be guix: only check for the macOS SDK once (fanquake)
Pull request description:
If we are building for both macOS HOSTS, there's no need to check and
print that the SDK exists two times.
Currently a Guix build for both HOSTS will print:
```bash
./contrib/guix/guix-build
Found macOS SDK at '/SDKs/Xcode-12.2-12B45b-extracted-SDK-with-libcxx-headers', using...
Found macOS SDK at '/SDKs/Xcode-12.2-12B45b-extracted-SDK-with-libcxx-headers', using...
Checking that we can connect to the guix-daemon...
```
ACKs for top commit:
laanwj:
Code review ACK e8023100be
achow101:
ACK e8023100be
Tree-SHA512: 7e9ee7793c5dc1eb485806ca3d613742397d2cc62525203a168cad1ec96aabfd4e63dc3f2e8d205bdb2f3c2079f731d75c5d162d55ff0d42a7e6f3d01d3a7db1
URLs may contain a query string (prefixed with '?') and this should be ignored when parsing
the data format.
To facilitate testing this functionality, ParseDataFormat has been made non-static.
As RetFormat is now exposed in a header, it is renamed to the more
understandable RESTResponseFormat
-BEGIN VERIFY SCRIPT-
s() { sed -i 's/RetFormat/RESTResponseFormat/g' $1; }
s src/rest.cpp
s src/rest.h
-END VERIFY SCRIPT-
c3296b21e4 build: Drop `double-conversion` from MSVC dependencies (Hennadii Stepanov)
7ff43e5372 ci: Invalidate vcpkg binary cache if dependencies changed (Hennadii Stepanov)
20b6c87117 build: Specify `zeromq` port explicitly for MSVC builds (Hennadii Stepanov)
Pull request description:
The current MSVC builds are broken due to the bug in the `zeromq` [port](https://github.com/microsoft/vcpkg/pull/22681#issuecomment-1061312320). From [IRC](https://bitcoin-irc.chaincode.com/bitcoin-core-dev/2022-03-08#787145):
> \<sipsorcery> Looks like it's a problem downloading the zeromq dependency from https://patch-diff.githubusercontent.com/raw/zeromq/libzmq/pull/4311.diff
> \<dhruv> sipsorcery: I'm definitely misunderstanding, i actually have no clue which file the CI is failing to download. I'll DM you more details.
> \<sipsorcery> It's saying the hash of the patch file has changed.
> \<dhruv> so we'd need to verify that the change is not malicious and then commit the new hash?
> \<sipsorcery> No that dependency is managed by the vcpkg repo. Seems they might be working on it https://github.com/microsoft/vcpkg/pull/22681#issuecomment-1061312320
> \<dhruv> ok, thanks
This PR fixes this issue with specifying the previous port version [explicitly](https://github.com/microsoft/vcpkg/blob/master/docs/users/versioning.md).
The current CI task does not fail due to the cached binaries.
---
The second commit makes vcpkg binary cache invalid if dependencies changed.
The third commit drops `double-conversion` from dependencies as Qt is configured as follows:
```
Configure summary:
Build type: win32-msvc (x86_64, CPU features: sse sse2)
Compiler: msvc 193131104
Configuration: sse2 aesni sse3 ssse3 sse4_1 sse4_2 avx avx2 avx512f avx512bw avx512cd avx512dq avx512er avx512ifma avx512pf avx512vbmi avx512vl compile_examples f16c largefile msvc_mp precompile_header rdrnd rdseed shani silent x86SimdAlways release c++11 c++14 c++17 c++1z concurrent no-pkg-config static static_runtime stl
Build options:
...
Qt Core:
DoubleConversion ....................... yes
Using system DoubleConversion ........ no
...
```
ACKs for top commit:
sipsorcery:
tACK c3296b21e4.
Tree-SHA512: 4d694a7d0930889a53eb6ee7a09929f6ffa3f078122b34abe6d75430769bb87c353f7c11146da53c3804e51d4bbfcbb7bc8453f525bcc432928d98eeb66ee35e
a84650ebd5 util: Fix ReadBinaryFile reading beyond maxsize (klementtan)
Pull request description:
Currently `ReadBinaryFile` will read beyond `maxsize` if `maxsize` is not a multiple of `128` (size of buffer)
This is due to `fread` being called with `count = 128` instead of `count = min(128, maxsize - retval.size()` at every iteration
The following unit test will fail:
```cpp
BOOST_AUTO_TEST_CASE(util_ReadWriteFile)
{
fs::path tmpfolder = m_args.GetDataDirBase();
fs::path tmpfile = tmpfolder / "read_binary.dat";
std::string expected_text(300,'c');
{
std::ofstream file{tmpfile};
file << expected_text;
}
{
// read half the contents in file
auto [valid, text] = ReadBinaryFile(tmpfile, expected_text.size() / 2);
BOOST_CHECK_EQUAL(text.size(), 150);
}
}
```
Error:
```
test/util_tests.cpp:2593: error: in "util_tests/util_ReadWriteFile": check text.size() == 150 has failed [256 != 150]
```
ACKs for top commit:
laanwj:
Code review ACK a84650ebd5
theStack:
Code-review ACK a84650ebd5
Tree-SHA512: 752eebe58bc2102dec199b6775f8c3304d899f0ce36d6a022a58e27b076ba945ccd572858b19137b769effd8c6de73a9277f641be24dfb17657fb7173ea0eda0
0189df1d31 build, mac: Include arch in codesignature tarball (Andrew Chow)
6e9308c6d4 guix: use latest signapple (Andrew Chow)
Pull request description:
Since we have two architectures for Mac binaries, having the architecture in the code signature tarball generated by `detached-sig-create.sh` allows us to avoid accidentally overwriting an existing code signature tarball during the code signing process.
ACKs for top commit:
fanquake:
ACK 0189df1d31
Tree-SHA512: 7e0d282e4ced1094f36f1d26ff6e18d53449561ab3a1a95ac69eb5ff3d7b33ee4bd8fad004884806064a29541c47f9e5879c2a1fd0f54453413245bdcf53c4c7
5b1aae12ca qt: Avoid crash on startup if int specified in settings.json (Ryan Ofsky)
84b0973e35 test: Add tests for GetArg methods / settings.json type coercion (Ryan Ofsky)
Pull request description:
Should probably add this change to 23.x as suggested by Luke https://github.com/bitcoin/bitcoin/issues/24457#issuecomment-1059825678. If settings like `prune` are added to `settings.json` in the future, it would be preferable for 23.x releases to respect the setting instead of crash.
---
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).
The fix is a one-line change in `ArgsManager::GetArg`. The rest of the PR just adds a regression test for the GUI and unit tests for ArgsManager::GetArg methods.
ACKs for top commit:
laanwj:
Code review ACK 5b1aae12ca
achow101:
ACK 5b1aae12ca
jonatack:
Code review ACK 5b1aae12ca
Tree-SHA512: 958991b4bead9b82a3879fdca0f8d6405e2a212b7c46cf356f078843a4f156e27fd75fc46e2013aa5159582ead06d343c1ed248d678b3e5bbd312f247e37894c
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
db27ac9354 tests: Ensure sorted/multi_a descriptors always generate different addrs (Andrew Chow)
Pull request description:
Sometimes the multi_a and sortedmulti_a descriptors will produce some of the same addresses in the tests. This causes the wallets to start generating addresses at a different index as they detect that one of the addresses is used. This subsequently causes a test failure.
To avoid this problem, use descriptors that will produce unique addresses by putting one of the multi_a in a different branch.
ACKs for top commit:
ajtowns:
ACK db27ac9354
theStack:
Tested ACK db27ac9354
Tree-SHA512: 0f57822bf4c7c79da304f092d7d43d6118e78a087cbeb0766fbbf634dc27911ae723d5d41350884d3b63a24d3b3817944f7e5fa534afb849161dd008a1e4a62f
