44c2452fd Merge bitcoin-core/secp256k1#1105: Don't export symbols in static libraries
6f6cab998 abi: Don't export symbols in static Windows libraries
485f608fa Merge bitcoin-core/secp256k1#1104: Fix the false positive of `SECP_64BIT_ASM_CHECK`
8b013fce5 Merge bitcoin-core/secp256k1#1056: Save negations in var-time group addition
7efc9835a Fix the false positive of `SECP_64BIT_ASM_CHECK`
2f984ffc4 Save negations in var-time group addition
git-subtree-dir: src/secp256k1
git-subtree-split: 44c2452fd387f7ca604ab42d73746e7d3a44d8a2
Achieve this by adding a MAIN_FUNCTION macro, consolidating the docs, and
introducing the macro across our distributed binaries.
Also update the docs to explain that anyone using binutils < 2.36 is
effected by this issue. Release builds are not, because they use binutils
2.37. Currently LTS Linux distros, like Ubuntu Focal, ship with 2.34.
https://packages.ubuntu.com/focal/binutils
fa7a711a30b707cbdee4435dd0a956bffb7aaccb test: Fix out-of-range port collisions (MacroFake)
Pull request description:
Otherwise the test will fail if two tests running in parallel use the same port. See https://github.com/bitcoin/bitcoin/pull/25096#discussion_r892558783 and https://github.com/bitcoin/bitcoin/pull/25312
ACKs for top commit:
dergoegge:
ACK fa7a711a30b707cbdee4435dd0a956bffb7aaccb - This gets rid of some rather arbitrary choices for ports in some of our functional tests that can cause port collisions across test runs, resulting in intermittent failures.
Tree-SHA512: ac73da8a498230b992ab12e1ee3c4ff3d868cd63c00d2c71537d156cb7c8f8be8598ec574646b17c5a44ae3ac5bb54bf29d300f054a36cec6f6ce8054a0da0a4
ce893c0497fc9b8ab9752153dfcc77c9f427545e doc: Update developer notes (Anthony Towns)
d2852917eecad6ab422a7b2c9892d351a7f0cc96 sync.h: Imply negative assertions when calling LOCK (Anthony Towns)
bba87c0553780eacf0317fbfec7330ea27aa02f8 scripted-diff: Convert global Mutexes to GlobalMutexes (Anthony Towns)
a559509a0b8cade27199740212d7b589f71a0e3b sync.h: Add GlobalMutex type (Anthony Towns)
be6aa72f9f8d50b6b5b19b319a74abe7ab4099ff qt/clientmodel: thread safety annotation for m_cached_tip_mutex (Anthony Towns)
f24bd45b37e1b2d19e5a053dbfefa30306c1d41a net_processing: thread safety annotation for m_tx_relay_mutex (Anthony Towns)
Pull request description:
This changes `LOCK(mutex)` for non-global, non-recursive mutexes to be annotated with the negative capability for the mutex it refers to, to prevent . clang applies negative capabilities recursively, so this helps avoid forgetting to annotate functions.
This can't reasonably be used for globals, because clang would require every function to be annotated with `EXCLUSIVE_LOCKS_REQUIRED(!g_mutex)` for each global mutex; so this introduces a trivial `GlobalMutex` subclass of `Mutex`, and reduces the annotations for both `GlobalMutex` to `LOCKS_EXCLUDED` which only catches trivial errors (eg (`LOCK(x); LOCK(x);`).
ACKs for top commit:
MarcoFalke:
review ACK ce893c0497fc9b8ab9752153dfcc77c9f427545e 🐦
hebasto:
ACK ce893c0497fc9b8ab9752153dfcc77c9f427545e
Tree-SHA512: 5c35e8c7677ce3d994a7e3774f4344adad496223a51b3a1d1d3b5f20684b2e1d5cff688eb3fbc8d33e1b9940dfa76e515f9434e21de6f3ce3c935e29a319f529
Code introduced in #15649 added usage of `timingsafe_bcmp()`, if
available, otherwise falling back to our own implementation. However
the relevant build system check was never added, so currently, we'll
always just use our implementation, as HAVE_TIMINGSAFE_BCMP will never
be defined.
Add the check for timingsafe_bcmp. Note that as far as I'm aware, it's
only available on OpenBSD.
ea54ba2f42f6d0b23570c665c2369f977bf55cf6 [test] Fix port collisions caused by p2p_getaddr_caching.py (dergoegge)
f9682e75ac184a62c7e29287882df34c25303033 [test_framework] Set PortSeed.n directly after initialising params (dergoegge)
Pull request description:
This PR fixes the issue mentioned [here](https://github.com/bitcoin/bitcoin/pull/25096#discussion_r892558783), to avoid port collisions between nodes spun up by the test framework.
Top commit has no ACKs.
Tree-SHA512: ec9159f0af90db636f7889d664c24e1430cf2bcb3c02a9ab2dcfe531b2a4d18f6e3a0f8ba73071bdf2f7db518df9d5d86a9cd06695e67644d20fe4515fac32b7
Rename `GenerateSelectSet()` to `GenerateWaitSockets()` and adapt it to
generate a wait data suitable for `Sock::WaitMany()`. Then call it from
`CConnman::SocketHandler()` and feed the generated data to
`Sock::WaitMany()`.
This way `CConnman::SocketHandler()` can be unit tested because
`Sock::WaitMany()` is mockable, so the usage of real sockets can be
avoided.
Resolves https://github.com/bitcoin/bitcoin/issues/21744
d575413fb8f8569803ed0050c15be003ef50951c doc: add `desig` to ignore-words (brunoerg)
c06cc41ddbac3611d5c7826a2a3f2cbae9aeea12 doc: fix typo in kernel/context.h (brunoerg)
Pull request description:
This PR fixes a typo in `kernel/context.h` (libary => library) and add `desig` to ignore-words since it's a valid word, see:
b9416c3847/src/net.cpp (L1105-L1117)
ACKs for top commit:
fanquake:
ACK d575413fb8f8569803ed0050c15be003ef50951c
Tree-SHA512: 2d548c737b8184d0243445c7503f3f68256ecb0970bd834d52de099de3cd8c8b9c140e2b77d55e2542fbd45b1d21cbdee639f5b2ef8138c37b8b72e5211029c3
It allows waiting concurrently on more than one socket. Being a
`virtual` `Sock` method it can be overriden by tests.
Will be used to replace `CConnman::SocketEvents()`.
8888bd43c100f9f0ca1122fcc896fb7b999d61c6 Remove redundant nLastTry check (MarcoFalke)
00001e57fe74c061aa9cbc72b07252335cb566e0 Remove redundant nTime checks (MarcoFalke)
Pull request description:
Split out from https://github.com/bitcoin/bitcoin/pull/24697 because it makes sense on its own.
ACKs for top commit:
dergoegge:
re-ACK 8888bd43c100f9f0ca1122fcc896fb7b999d61c6
naumenkogs:
utACK 8888bd43c100f9f0ca1122fcc896fb7b999d61c6
Tree-SHA512: 32c6cde1c71e943c76b7991c2c24caf29ae467ab4ea2d758483a0cee64625190d1a833b468e8eab1f834beeb2c365af96552c14b05270f08cf63790e0707581d
fa74b63c01db412f6a4378cb669d89496a89d02e test: Fix wait_for_debug_log UnicodeDecodeError (MacroFake)
Pull request description:
Fix the intermittent `UnicodeDecodeError` when the debug log is truncated on an (multi-byte) unicode character by treating everything as bytes.
Also, remove the `ignore_case` option and the`re.search+re.escape` wrap. All of this is unused and doesn't exist on raw byte strings.
Fixes https://github.com/bitcoin/bitcoin/issues/24575
ACKs for top commit:
jonatack:
ACK fa74b63c01db412f6a4378cb669d89496a89d02e
brunoerg:
ACK fa74b63c01db412f6a4378cb669d89496a89d02e
Tree-SHA512: c67c9355073e784fa8d9d48b8e79ff0c98f5ae9cd4d704ad12a76d2604733946054bc74b8ab346aa2184db23d740b85c8c13eb892d76cba92e42ebfd73f2f1bf
06e18e0b53ed34933ecd7f3976a508be72f687aa build: use BOOST_MULTI_INDEX_ENABLE_SAFE_MODE when debugging (fanquake)
Pull request description:
Use of this macro enables precondition checks for iterators and functions of the library. It's use is recommended in debug builds. See https://www.boost.org/doc/libs/1_78_0/libs/multi_index/doc/tutorial/debug.html for more info.
There is also a `BOOST_MULTI_INDEX_ENABLE_INVARIANT_CHECKING` macro:
> When this mode is in effect, all public functions of Boost.MultiIndex will perform post-execution tests aimed at ensuring that the basic internal invariants of the data structures managed are preserved.
ACKs for top commit:
laanwj:
Concept and code review ACK 06e18e0b53ed34933ecd7f3976a508be72f687aa
Tree-SHA512: 7ee489eccda81c7dbca9210af6d3007d5b2c704b645139d2714c077af157789dd9478c29d0d212e210e96686ea83713aaf3d458e879122b3cde64f3e3e3789d2
Here we update only the log messages that manually print a category.
In upcoming commits, LogPrintCategory will likely be used in many
other cases, such as to replace `LogPrintf` where it makes sense.
292828cd7744ec7eadede4ad54aa2117087c5435 [test] Test addr cache for multiple onion binds (dergoegge)
3382905befd23364989d941038bf7b1530fea0dc [net] Seed addr cache randomizer with port from binding address (dergoegge)
f10e80b6e4fbc151abbf1c20fbdcc3581d3688f0 [net] Use ConnectedThroughNetwork() instead of GetNetwork() to seed addr cache randomizer (dergoegge)
Pull request description:
The addr cache id randomizer is currently supposed to be seeded with the network of the inbound connection and the local socket (only the address is used not the port): a8098f2cef/src/net.cpp (L2800-L2804)
For inbound onion connections `CNode::addr.GetNetwork()` returns `NET_UNROUTABLE` and `CNode::addrBind` is set to `127.0.0.1:<onion bind port>`. This results in the same addr cache for all inbound connections on 127.0.0.1 binds.
To avoid the same addr cache across all onion and other 127.0.0.1 binds, we should seed the addr cache randomizer with the correct network for inbound onion connections (using `CNode::ConnectedThroughNetwork()`) as well as the port of `CNode::addrBind`.
ACKs for top commit:
sipa:
utACK 292828cd7744ec7eadede4ad54aa2117087c5435
mzumsande:
Code Review ACK 292828cd7744ec7eadede4ad54aa2117087c5435
naumenkogs:
utACK 292828cd7744ec7eadede4ad54aa2117087c5435
Tree-SHA512: d0be13bab6bc121c2926d4b168687f6c2ed4ce0c9dd19be71eb4886adeba8afc3daacdc4e232a0ba3b03a89d69b618abc5595b69abd1ad0c476d825bc6ea1f9f
Now that we use GCC 10 for release builds, we no-longer need to
pass-Wl,-z,noexecstack to get a non-executable stack in RISC-V binaries.
This was originally removed in #21036, but then re-added in #21799, when
we reverted to using GCC 8.
433b52569417674f84c2b1d449037701814420c4 Add LogPrintLevel to lint-format-strings, drop LogPrint-vs-LogPrintf section in dev notes (Jon Atack)
Pull request description:
added by #7003 in 2015, as that potential issue would now be caught by the `test/lint/lint-format-strings.py` script run by the CI.
ACKs for top commit:
MarcoFalke:
cr ACK 433b52569417674f84c2b1d449037701814420c4
w0xlt:
ACK 433b525694
Tree-SHA512: 91a2ac76689ed4f1f638e07c16d2ec8952fb013cc8bb896780fbd9333abd084281ce99afdc9de715d07a9abb4dce5dd67edf5e347aff466c6ef339ccc4158679
57fb37c27599fc865f20b42a27bb9c227f384de3 wallet: CommitTransaction, remove extra wtx lookup and add exception for a possible db write error. (furszy)
Pull request description:
Two points for `CWallet::CommitTransaction`:
1) The extra wtx lookup:
As we are calling to `AddToWallet` first, which returns the recently added/updated wtx pointer, there is no need to look up the wtx again few lines later. We can just use it.
2) The db write error:
`AddToWallet` can only return a nullptr if the db write fails, which inside `CommitTransaction` translates to an exception throw cause. We expect everywhere that `CommitTransaction` always succeed.
------------------------------------------------
Extra note:
This finding generated another working path for me :)
It starts with the following question: why are we returning a nullptr from `AddToWallet` if the db write failed without removing the recently added transaction from the wallet's map?..
Can led to a wallet invalid state where the inputs of this new transaction are not marked dirty, while the transaction that spends them still exist on the in-memory wallet tx map.
-- I'm writing it here to gather some feedback first and not forget it, will create a follow-up PR in the coming days 🚜 --
ACKs for top commit:
achow101:
ACK 57fb37c27599fc865f20b42a27bb9c227f384de3
jonatack:
ACK 57fb37c
ryanofsky:
Code review ACK 57fb37c27599fc865f20b42a27bb9c227f384de3. Seems like a clear improvement. Better to fail earlier with a better error message if the failure is going to happen anyway
Tree-SHA512: 80e59c01852cfbbc70a5de1a1c2c59b5e572f9eaa08c2175112cb515256e63fa04c7942f92a513b620d6b06e66392029ebe8902287c456efdbee58a7a5ae42da
687addaf136356e0f3698d6345c92d875e0a3362 test: add BIP-125 rule 5 testcase with default mempool (James O'Beirne)
6120e8e2871fecfec5ab3099c97e13951e062a4d test: allow passing sequence through create_self_transfer_multi (James O'Beirne)
Pull request description:
Currently, we only test rule 5 of BIP-125 (replacement transactions cannot evict more than 100 transactions) by changing default mempool parameters to allow for more descendants. The current test works on a single transaction graph that has over 100 descendants.
This patch adds a test to exercise rule 5 using the default mempool parameters. The case is a little more sophisticated: instead of working on a single transaction graph, it uses a replacement transaction to "unite" several UTXOs which join independent transaction graphs. The total number of transactions in these graphs sum to more than the max allowable replacement.
I think the difference in transaction topology makes this a worthwhile testcase to have, setting aside the fact that this testcase works without having to use atypical mempool params.
See also: [relevant discussion from IRC](https://www.erisian.com.au/bitcoin-core-dev/log-2022-05-27.html#l-126)
ACKs for top commit:
laanwj:
Code review ACK 687addaf136356e0f3698d6345c92d875e0a3362
LarryRuane:
ACK 687addaf136356e0f3698d6345c92d875e0a3362
Tree-SHA512: e589aeaf9d6f137d546b7809f8795d6f6043d87b15e97c2efe85b42ce8b49d977ee7d79440c542ca4b0b5ca2de527488029841a1ffc0d96c5771897df4b3f324
that was added in 2015 by commit b8c06ef40 in PR 7003, as that potential issue
would now be caught by the test/lint/lint-format-strings.py script run by the CI
d40550d725fa7afd6916510164bbaa95e3ec0a5f scripted-diff: remove duplicate categories from LogPrint output (Jon Atack)
Pull request description:
This is the first commit from #25203.
- Scripted-diff: de-duplicate logging category output for the tor, i2p, net, zmq, and prune messages (e.g. where I found duplicates), as these category prefixes are now printed automatically since #24464
examples before
```
[tor] tor: Successfully connected!
[i2p] I2P: Creating SAM session with 127.0.0.1:7656
[zmq] zmq: Initialize notification interface
[net] net: enabling extra block-relay-only peers
```
after
```
[tor] Successfully connected!
[i2p] Creating SAM session with 127.0.0.1:7656
[zmq] Initialize notification interface
[net] enabling extra block-relay-only peers
```
ACKs for top commit:
klementtan:
crACK d40550d725fa7afd6916510164bbaa95e3ec0a5f
MarcoFalke:
cr ACK d40550d725fa7afd6916510164bbaa95e3ec0a5f
Tree-SHA512: 63b799f2f899f0597981dd1acb91ef4439cd00b257a9eb19d67c4ce2c4dc72a95ac5761cb78f2a19090a10be74f23ea1db6929ed942ba0d008b4be563f0d5e7e
fa4068b4e2192f168bb120624eca5735f0dadf6f Move minRelayTxFee to policy/settings (MacroFake)
Pull request description:
Seems a bit confusing to put policy stuff into validation, so fix that.
Also fix includes via `iwyu`.
ACKs for top commit:
ariard:
ACK fa4068b, the includes move compiles well locally.
ryanofsky:
Code review ACK fa4068b4e2192f168bb120624eca5735f0dadf6f. Make sense to move the global variable to policy/settings and the default constant to policy/policy. Ariard points out other constants that could be moved, which seems fine, but it seems like moving the global variable to be with other related global variables is more significant.
Tree-SHA512: adf9619002610d1877f3aef0a9e6115fc4c2ad64135a3e5100824c650b560c47f47ac28894c6214a50a7888355252a9f6f7cec98c23a771a1964160ef1ca77de
f26a496dfd0a7ce3833a10075027d7d5b0345e32 test: clean up all-lint.py (Martin Leitner-Ankerl)
64d72c4c8734b9dd45cb61cb2c2baf98766b0163 test: rename lint-all.py to all-lint.py (Martin Leitner-Ankerl)
Pull request description:
When running `./test/lint/lint-all.py`, the script runs all tests but
also calls itself because the comparison with `__file__` doesn't work.
Comparing resolved paths gives reliable comparison, and lint-all.py doesn't call itself any more
ACKs for top commit:
laanwj:
Code review ACK f26a496dfd0a7ce3833a10075027d7d5b0345e32
Tree-SHA512: b44abdd685f7b48a6a9f48e96d97138b635c31c1c7ab543cb5636b5f49690ccd56fa6fec01ae7fcc16af01a613372ee77632f70c32059919b373aa8051953791
e593ae07c4fb41a26c95dbd03301607fc5b4d5e2 Bugfix: RPC/blockchain: pruneblockchain: Return the height of the actual last pruned block (Luke Dashjr)
Pull request description:
From 0.14 (2017 Mar) until before 0.19 (2019 Nov), the height of the last block pruned was returned, subject to a bug if there were blocks left unpruned due to sharing files with later blocks.
In #15991, this was "fixed" to the current implementation, introducing a new bug: now, it returns the first *unpruned* block.
Since the user provides the parameter as a block to include in pruning, it makes more sense to fix the behaviour to match the documentation.
~~(Additionally, the description of "pruneheight" in getblockchaininfo is fixed to be technically correct)~~
ACKs for top commit:
fjahr:
utACK e593ae07c4fb41a26c95dbd03301607fc5b4d5e2
ryanofsky:
Code review ACK e593ae07c4fb41a26c95dbd03301607fc5b4d5e2. Just rebased since last review. Maybe some of the original reviewers of #15991 will want to take a look at this to correct the mistake that was introduced there!
Tree-SHA512: c2d511df80682d57260aae8af1665f9d7eaed16448f185f4c9f23c78fa9b8289a02053da7a0b83643fef57610d601ea63b59ff39661a51f4827f1eb27cc30594
...also adjust callers
Changes:
- In BlockAssembler::CreateNewBlock, we now only lock m_mempool->cs and
call addPackageTxs if m_mempool is not nullptr
- BlockAssembler::addPackageTxs now takes in a mempool reference, and is
annotated to require that mempool's lock.
- In TestChain100Setup::CreateBlock and generateblock, don't construct
an empty mempool, just pass in a nullptr for mempool
3a9b9bb38e653c8ff7220b9af6e337a90c2c22dc test: ensure createmultisig and addmultisigaddress are not returning any warning for expected cases (brunoerg)
eaf6f630c0190c634b5f1c85f749437f4209cc36 rpc: fix inappropriate warning for address type p2sh-segwit in createmultisig and addmultisigaddress (brunoerg)
Pull request description:
Fixes#25127
If there are any uncompressed keys when calling `AddAndGetMultisigDestination`, it will just default to a legacy address regardless of the chosen `address_type`. So, #23113 added a warnings field which will warn the user why their address format is different.
However, when creating a multisig (p2sh-segwit), it is returning an inappropriate warning, because when getting the output type from destination (`OutputTypeFromDestination`), it returns `ScriptHash` for both legacy and `P2SH_SEGWIT`. So, since `P2SH_SEGWIT` is different from `ScriptHash`, it returns the warning:
192d639a6b/src/rpc/output_script.cpp (L166-L169)
So, to avoid this mistake I changed `OutputTypeFromDestination` to `descriptor->GetOutputType()` to get the appropriate output type.
ACKs for top commit:
jonatack:
ACK 3a9b9bb38e653c8ff7220b9af6e337a90c2c22dc
laanwj:
Code review ACK 3a9b9bb38e653c8ff7220b9af6e337a90c2c22dc
Tree-SHA512: 49f717479c2b8906277e7591ddd4747f7961c2d5c77494b5124045de9036a4277d46b9ad99279d51f0c4484284c445f1e1d3c55c49bbf0716741bad426a89369
c47944f4e93204e13686280da86efffee4ba6eb2 ci: Reuse some configure options in "ARM64 Android APK" task (Hennadii Stepanov)
7739438811adef67766aa845122633ff7f594e35 ci, android: Update NDK up to r23c (Hennadii Stepanov)
ca0c3e50779771832dd8db419940fde877d9b691 ci, android: Update Command-line Tools from 2.1 up to 7.0 (Hennadii Stepanov)
8790da3c1e9006ca9c1937595cb0f55889c1c026 ci: Drop unneeded packages in "ARM64 Android APK" task (Hennadii Stepanov)
Pull request description:
This PR improves the "ARM64 Android APK" CI task in the following ways:
- dropped packages that are not required to be installed
- updated Android Command-line Tools and Android NDK to make the CI environment closer to the default one, which is provided by Android Studio
ACKs for top commit:
icota:
utACK c47944f4e93204e13686280da86efffee4ba6eb2
Tree-SHA512: 45f5aba41007a502ae90333272370fd559c48a27d573896c449b3e436c5cf2b6440408381e4d20eb53104426ade26d3a9014c09dcdf3257ec897a537095efa4f
d1b7bcbca2be4eae3bf7572a8343a3bd69ce61a9 qt: Drop no longer supported Android architecture (Hennadii Stepanov)
Pull request description:
The `i686-linux-android` arch support has been dropped since bitcoin/bitcoin#23744.
ACKs for top commit:
katesalazar:
ACK d1b7bcbca2be4eae3bf7572a8343a3bd69ce61a9
icota:
utACK d1b7bcbca2
prusnak:
Approach ACK d1b7bcbca2be4eae3bf7572a8343a3bd69ce61a9
Tree-SHA512: 13689ec8c63c92b9a52a3c25edc35536b8e51ff583f57c45b168515f928d020d6bb85d03db9efd8d5efd57b944dfd313a89f5ff8a52f99982ccc8d9671f6e7a9
