1869310f3cfa4ab26b5090d8a4002eefdc84870e refactor: remove unused param from legacy pubkey (Bushstar)
Pull request description:
Unused param present in legacy pubkey manager interface. This param will not be used and should be removed to prevent unintended usage.
ACKs for top commit:
Sjors:
ACK 1869310f3cfa4ab26b5090d8a4002eefdc84870e
furszy:
ACK 1869310f3cfa4ab26b5090d8a4002eefdc84870e
Tree-SHA512: 0fb41fc8f481f859262f2e8e9a93c990c1b4637e74fd9191ccc0b3c523d0e7d94217a3074bb357276e1941a10d29326f850f9b27eccc1eca57cf6b549353400c
Also, move the burden of checking for a timeout to the client and
disable the timeout on the server. This should avoid intermittent issues
in slow tests (for example mining_getblocktemplate_longpoll.py, or
feature_dbcrash.py), or possibly when the server is running slow (for
example in valgrind). There shouldn't be any downside in tests caused
by a high rpcservertimeout.
fafe3a8e38bdb6d12b2654b066b8036ba3e9455d ci: Remove second user account (MarcoFalke)
Pull request description:
The rationale for the second (nonroot) account no longer applies. See also https://github.com/bitcoin/bitcoin/pull/27333#discussion_r1148898438
ACKs for top commit:
josibake:
utACK fafe3a8e38
Tree-SHA512: 4464e50bc6499e012ff33c591f20c5989b99e223c476cefea669aedb79a0e731734db9469d4c7d800642191b43b57ea32cda187093395f100dec719fea17724d
Taking cs_main is no longer necessary since we moved
`m_recently_announced_invs` to `Peer` and `mapRelay` is actually only
accessed from the message processing thread.
71b3e9b0ade9680f6847e93785225c5927929336 sanitizers: remove GetRNGState lsan suppression (fanquake)
Pull request description:
I am no-longer seeing this, testing with the native_asan job over `x86_64` (Ubuntu 22.04) and `aarch64` (Fedora 37).
Can anyone recreate the false-positive?
ACKs for top commit:
MarcoFalke:
lgtm ACK 71b3e9b0ade9680f6847e93785225c5927929336
hebasto:
ACK 71b3e9b0ade9680f6847e93785225c5927929336, tested on Ubuntu 22.04 x86_64.
Tree-SHA512: 63020327d61acd6c94c6c278c9c4d72aedc10253fa172bcf9353bcad4c28d068bee824969eb3ce92152244831df8fe92cffae536453c8073a4fda74dfdfbcefa
Rewrite the same algo instead of reusing BlockAssembler because we have
a few extra requirements that would make the changes invasive and
difficult to review:
- Only operate on the relevant transactions rather than full mempool
- Remove transactions that will be replaced so they can't bump their ancestors
- Don't hold mempool lock outside of the constructor
- Skip things like max block weight and IsFinalTx
- Additionally calculate fees to bump remaining ancestor packages to target feerate
Co-authored-by: Murch <murch@murch.one>
d0e571ebb187d7c4c2821f1334cb2dd4222dd8ce guix: use python-minimal (3.9) (fanquake)
Pull request description:
This further minifies the Guix release build environment.
ACKs for top commit:
TheCharlatan:
ACK d0e571ebb187
hebasto:
ACK d0e571ebb187d7c4c2821f1334cb2dd4222dd8ce
Tree-SHA512: 0a8aa9ae861107f106c3b9c41f78ffbaf0e71e3c61f6d96e5c82415b4570b8ac85d6578d37cd0df0ec315c1c9f35fc90b281f139271ccfd15a1495ba76166789
b5ef1419ece77db77cc196eb541f99a72360a607 ci: cleanup of CI_EXEC & CI_EXEC_ROOT (refs #27321) (Vasil Stoyanov)
Pull request description:
Basically it removes the above-mentioned env-vars as per MarcoFalke's instructions. The only deviation from the plan laid out there was that I double-quoted the last instance of $ANDROID_HOME for the sake of consistency and future-proofing and the rest of the non-quoted vars due to lint failing the build.
Fixes #27321.
ACKs for top commit:
josibake:
ACK b5ef1419ec
hernanmarino:
untested ACK b5ef1419ece77db77cc196eb541f99a72360a607. LGTM
Tree-SHA512: a79776bf64a2fa8b38195cc84445e171fd689f156aac5a1e5d39040300567eb9f4c2ebd00fbf3fa0e55b68793f8f752d94f7d817f6097ed9dd3a8ea57651b981
e669833943bda13b2840a174dc8e45194187fc8e test: dedup package limit checks via decorator in mempool_package_limits.py (Sebastian Falbesoner)
72f25e238c1f791f9fd3018152d76f9127b745e2 test: refactor: use Satoshis for fees in mempool_package_limits.py (Sebastian Falbesoner)
Pull request description:
The subtests in the functional test mempool_package_limits.py all follow the same pattern:
1. first, check that the mempool is currently empty
2. create and submit certain single txs to the mempool, prepare list of hex transactions
3. check that `testmempoolaccept` on the package hex fails with a "package-mempool-limits" error on each tx result
4. after mining a block, check that submitting the package succeeds
Note that steps 1,3,4 are identical for each of the subtests and only step 2 varies, so this might be a nice opportunity to deduplicate code by using a newly introduced decorator which executes the necessary before and after the essential part of the subtest. This also makes it easier to add new subtests without having to copy-paste those parts once again.
In addition, the first commit switches the fee unit from BTC to Satoshis, which allows to get rid of some imports (`COIN` and `Decimal`) and a comment for the `test_desc_size_limits` subtest is fixed (s/25KvB/21KvB/).
ACKs for top commit:
ismaelsadeeq:
ACK e669833943bda13b2840a174dc8e45194187fc8e
glozow:
utACK e669833943bda13b2840a174dc8e45194187fc8e
Tree-SHA512: 84a85e739de7387391c13bd46aeb015a74302ea7c6f0ca3d4e2b1b487d38df390dc118eb5b1c11d3e4206bff316a4dab60ef6b25d8feced672345d4e36ffd205
3a060ae7b67cc28fc60cf28cbc518fa1df24f545 scripted-diff: Rename nUnconnectingHeaders and fPreferHeaders (dergoegge)
279c53d7e4e920d3d09da74689761a2961b1a636 [net processing] Move m_recently_announced_invs from CNodeState to Peer (dergoegge)
938a8e2566bf2f819adeb75a08c0c8ba9a85fe1c [net processing] Annotate m_recently_announced_invs as guarded by g_msgproc_mutex (dergoegge)
8a2cb1f749ade31bb0655567cd0062db8a7164e5 [net processing] Move fPreferHeaders from CNodeState to Peer (dergoegge)
3605011e79d0d43ff532a2c3cc9cad4cc48eb06c [net processing] Annotate fPreferHeaders as guarded by g_msgproc_mutex (dergoegge)
4b84e502f5ba69d650191ce8f2e20b86c1d427ed [net processing] Move m_headers_sync_timeout from CNodeState to Peer (dergoegge)
689b747fc34a5670a24bb5c4b1986dd8670b55ef [net processing] Annotate m_headers_sync_timeout as guarded by g_msgproc_mutex (dergoegge)
d8c0d1c345546c6066274edbf3bf4fb5d6f75419 [net processing] Move nUnconnectingHeaders from CNodeState to Peer (dergoegge)
5f80d8d1eec14253407ceed1a813619d3e5a0fe4 [net processing] Annotate nUnconnectingHeaders as guarded by g_msgproc_mutex (dergoegge)
1d87137227ee125018006935d55164a72265c935 [validation] Annotate ChainstateManager::m_best_header as guarded by cs_main (dergoegge)
Pull request description:
`nUnconnectingHeaders`, `m_headers_sync_timeout`, `fPreferHeaders` and `m_recently_announced_headers` are currently all `CNodeState` members even though they are only ever accessed from the message processing thread (therefore sufficiently guarded exclusively by `g_msgproc_mutex`). `CNodeState` exists purely to hold validation-specific state guarded by `cs_main` that is accessed by multiple threads.
This PR adds thread-safety annotations for the above mentioned `CNodeState` members and moves them to `Peer`.
ACKs for top commit:
glozow:
code review ACK 3a060ae7b67cc28fc60cf28cbc518fa1df24f545, as in I am convinced these members shouldn't be guarded by cs_main and belong in Peer/TxRelay. clang checked the annotations for me.
hebasto:
ACK 3a060ae7b67cc28fc60cf28cbc518fa1df24f545
Tree-SHA512: 2db27c03f2c6ed36ad7dfbb4f862eeed3c3e57f845cf8abb9e7cada36f976257311892020bbcff513fbe662a881c93270e3a126946ceb0c3f94213b546bcaa81
a634c288c34627eb4ea6f919c99339256f490ede ci: use LLVM/clang-16 in native_fuzz (ASAN) job (fanquake)
Pull request description:
Similar to #27298.
ACKs for top commit:
dergoegge:
utACK a634c288c34627eb4ea6f919c99339256f490ede
Tree-SHA512: 7a2625a3ac83710063d941dcbca42431b3b79a1380872fd2c566c0ab3041d8123d7dcddeb8a4972efd0ef6496b15bbe0b39b6d2de84df81fcdd8d68e1248fbc5
ea7ec7808745805c0a18513d7da271dedb2de3f1 refactor: Drop no longer used `CNetMsgMaker` instances (Hennadii Stepanov)
Pull request description:
The removed lines have been unused since the abf5d16c24 commit from https://github.com/bitcoin/bitcoin/pull/25454.
ACKs for top commit:
dergoegge:
utACK ea7ec7808745805c0a18513d7da271dedb2de3f1
Sjors:
ACK ea7ec7808745805c0a18513d7da271dedb2de3f1
TheCharlatan:
ACK ea7ec7808745805c0a18513d7da271dedb2de3f1
Tree-SHA512: 9a2a9ff3f124b68a8cd20a637e90885096996c3aa354a4d8adbec98f5761e9e826c1c064ccd90aaf6d72beac61dd9e22c8b76d089e18bba6e0ad51e59a9c7df8
4133c8104f522c403c55d26bd03436a8149ff106 guix: use gcc tool wrappers (fanquake)
Pull request description:
This way, correct `--plugin` arguments are passed through.
This is a prerequisite for LTO (see #25391). Split out, to try move things along, as this change is isolated, and should be straight-forward.
ACKs for top commit:
TheCharlatan:
ACK [4133c81](4133c8104f)
hebasto:
ACK 4133c8104f522c403c55d26bd03436a8149ff106
Tree-SHA512: 4311a72a613cf027bd4490caa29604c985ed455589acd972285f13cbdf4806d2184a4dc6f20cb6f47c3fa751d58bfd0bacc257b87d4a804bf5ecf5b240e4a757
e47ce42f670fc43859c157766b342509ab5916f9 refactor: use address_to_scriptpubkey to retrieve addresses scriptpubkey (ismaelsadeeq)
4142d19d741d6432ba95f3452f0d949941d89d5c refactor: move address_to_scriptpubkey to address.py (ismaelsadeeq)
Pull request description:
PR #27269 enables the function address_to_scriptpubkey() to decode all address types and return their corresponding scriptpubkeys. As a result, there is no longer any need to call getaddressinfo or validateaddress RPCs in order to retrieve an address scriptpubkey, as explained in the comments on this pull request (see https://github.com/bitcoin/bitcoin/pull/27269#pullrequestreview-1353681933 and https://github.com/bitcoin/bitcoin/pull/27269#issuecomment-1481016118).
Instead of using RPC calls, this update replaces the process of obtaining an address scriptPubkey with the address_to_scriptpubkey method, resulting in improved performance for functional tests.
ACKs for top commit:
josibake:
re-ACK e47ce42f67
theStack:
ACK e47ce42f670fc43859c157766b342509ab5916f9 🌱
Tree-SHA512: 05285349a7d5ce7097b8f2582e573a5135c6deef85ea9936f68f6ce94e9ebb1d84d94f7fc7e5ed833a698e01585addd80deb52e6338f8aee985bf14db45417d2
In addition to verifying the published releases with the `pub` command,
the verifybinaries script is updated to take a `bin` command where the
user specifies the local files, sums, and sigs to verify.
This commit adds the functionality necessary to transition from
doing binary verification on the basis of a single signature to
requiring a minimum threshold of trusted signatures.
A signature can appear as "good" from GPG output, but it may not come
from an identity the user trusts. We call these "good, untrusted"
signatures.
We report bad signatures but do not necessarily fail in their presence,
since a bad signature might coexist with enough good, trusted signatures
to fulfill our criteria.
If "--import-keys" is enabled, we will prompt the user to
optionally try to retrieve unknown keys. Marking them as trusted locally
is a WIP, but keys which are retrieved successfully and appear on the
builder-keys list will immediately count as being useful towards
fulfilling the threshold.
Logging is improved and an option to output JSON that summarizes the
whole sum signature and binary verification processes has been added.
Co-authored-by: Russ Yanofsky <russ@yanofsky.org>
Co-authored-by: willcl-ark <will8clark@gmail.com>
